INVITED & CONTRIBUTED Through EasyChair

7th International Conference on Data Science and SDGs: Challenges, Opportunities and Realities

EC-001 Selecting Covariance Structure to Analyze Longitudinal Data: A Study to Model the Body Mass Index of Primary School Going Children in Bangladesh Mohammad Ohid Ullah1*, and Mst. Farzana Akter2 1Professor, Department of Statistics, Shahjalal University of Science and Technology, Sylhet, Bangladesh Email: [email protected] 2MSc Thesis student, Department of Statistics, Shahjalal University of Science and Technology, Sylhet, Bangladesh, Email: [email protected]

Background In longitudinal study, the data are collected from the same subject over time and hence data are correlated. To analyze these kinds of data, selecting a better covariance structure is very important to get the better results. Therefore, we aimed to select better covariance structure to model the body mass index (BMI) of primary school going children in Bangladesh.

Methods To do this study, at first we conducted a longitudinal study, we built a cohort of100primary school going children in Sylhet city, Bangladesh. We collected the information from the same children at the initial time

(T0), after six month (T6),after twelve month (T12) and after 18 months (T18).Linear Mixed Model (LMM) was applied for selecting a better covariance structure and then to model the body mass index. Results To find out a better covariance structure, we used Diagonal, Unstructured (UN), Auto Regressive order 1 (AR1) and Compound Symmetry (CS) covariance structures in the collected longitudinal data. Observing all the criteria, it is found that the covariance structure ‘CS’ (compound symmetry) gives better results of LMM. Finally using the CS covariance structure, we observed thatover time BMI of male students’ are comparatively less than female students’ (-0.040, p=0.034). Conclusions Taken together, we may conclude that Compound Symmetry (CS) gives better output to model the body mass index of primary school going children. As female students are getting more obese, in addition, today’s female children are the mothers of the future; therefore, parents should give concentration on female children to reduce their weight. This study may be useful for researchers in public health sectors to select a proper covariance structure to analyze their longitudinal data. Keywords: Compound Symmetry, Obesity, Longitudinal study, Linear Mixed Model, Public Health.

EC-002 Does an empirical relationship exist between financial development and economic growth in SAARC countries? Md Abdullah Al Mamun1, Ranjan Kumar Mitra2 and Md Abdul Wadud3 1 Associate Professor, Department of Folklore, Faculty of Social Sciences, University of Rajshahi Email: [email protected] 2 Associate Professor, Department of Accounting & Information Systems, University of Dhaka Email: [email protected] 3 Professor, Department of Economics, University of Rajshahi Email: [email protected]

Economic theories suggest that a sound and efficient financial system promotes economic growth by fostering efficient allocation of resources. This paper aims to investigate the theoretical prediction about the relationship between financial development and economic growth in SAARC countries. Generalized Method of Moment (GMM) has been applied on panel data of five SAARC countries covering the period from 1984 to 2016 to test the hypothesized relationship. As an empirical proxy for economic growth, the study uses annual percentage growth rate of GDP per capita. We employ four different indicators of financial development based on domestic credit provided by the financial sector institutions of respective economies. Result reveals notable positive effect of financial development on economic growth in SAARC countries, after controlling the time-constant country-specific effect or effects of potential biases induced by

103

simultaneity. Thus, findings are consistent with the theories that predict a significant positive role of financial development in the process of economic growth of a country. Keywords: Financial development, Economic growth, GMM method, SAARC economy

EC-003 Undernutrition among Bhumij preschool children of West Bengal, India Biswajit Mahapatra1, 2, Soma Pal1, Kaushik Bose1* and Goutam Sahoo1 1Department of Anthropology, Vidyasagar University, Midnapore-721102, West Bengal, India. 2Anthropological Survey of India, Head Office, 27, Jawaharlal Nehru Road, Kolkata-700016, West Bengal, India.

Poor nutritional status among tribal preschool children observed in different regions of West Bengal as well as India. Nutritional status of tribal preschool children especially Bhumij children of West Bengal has not been investigated adequately. The present community based cross-sectional study was undertaken to determine the prevalence of undernutrition using stunting, underweight and wasting among Bhumij preschool children of West Bengal, India. A total of 245 (129 boys; 116 girls) Bhumij children aged from 1 to 5 years were studied from eleven villages of Keshiary, Dantan-I and Nayagram block of Paschim Medinipur and Jhargram district of West Bengal. Length/height and weight of children was measured according to standard procedures. Date of birth was collected from birth certificates or health immunization cards of children. WHO Child Growth Standards (2006) was used to evaluate the nutritional status of Bhumij preschool children. Results revealed that there were no significant sex differences in mean height. Only 5 years (t=2.26 and p<0.05) Bhumij boys were significantly heavier than girls counterpart. Significant age differences in mean height [Height: F=213.66, p<0.001 (boys) and F=223.0, p<0.001 (girls)] and weight [Weight: F=116.95, p<0.001 (boys) and F=88.77, p<0.001 (girls)] was observed for both sexes. Overall prevalence of stunting, underweight and wasting were 45.7%, 24.9% and 9.8%, respectively. The age combined prevalence of undernutrition among boys were higher than girls (stunting: 51.2% vs 39.7%; underweight: 27.9% vs 21.6% and wasting: 14.0% vs 5.2%). Very high prevalence of stunting indicates critical situation of these communities. There is an urgent need to improve health care services to the tribal population and tribal children. Short Running Title: Prevalence of undernutrition among Bhumij preschool children. KEY WORDS: West Bengal; Bhumij; Undernutrition; Pre-school children.

EC-004 Investigating the role of climate change on dengue emergence and transmission in Bangladesh Md. Azizul Baten PhD Professor, Department of Statistics, Shahjalal University of Science and Technology Sylhet-3114, Bangladesh Email: [email protected]; [email protected]

Dengue is a viral infectious disease that is transmitted by mosquitoes, increased over time, the solution is really needed to control the dengue fever since it causes to mortality. The main objective of this study is to examine the impact of climate change on dengue fever and to establish the relationship of dengue fever occurrences as a linear function of health care system performance and a set of climate factors using Poisson regression model. The coefficient of maximum temperature with (0.8419), minimum temperature (-0.8539), rainfall (0.0242), humidity (0.4261), general health efficiency score estimated from Battese and Coelli (1992) with (-110.403) are significant while the coefficient of maximum temperature with (0.7368), minimum temperature (-0.3848), rainfall (0.0232), humidity (0.1939), general health efficiency score estimated from Battese and Coelli (1995) with (-132.112) are found highly significant for dengue transmission using Poisson regression model. The climate factors such as maximum temperature, rainfall and humidity had shown positive impact on dengue transmission while the minimum temperature, and health efficiency had shown negative impact on dengue transmission. This study provides the message to health policy makers in exploring the potential strategies in order to reduce the existing burden of the increase in dengue transmission in Bangladesh. Keywords: Dengue, Climate Change, Health Performance, Stochastic Frontier Analysis, Poisson Regression Analysis, Bangladesh

104

EC-005 An Efficient Estimation Procedure of Population Median in Two-Occasion Successive Sampling A. K. Singh Department of Statistics, Sri Venkateswara College, University of Delhi, India [email protected]

The objective of this paper is to estimate the finite population median using information on auxiliary variable in two-occasion successive sampling. General class of estimators have been proposed. Properties of the proposed estimators have been studied including optimum replacement policy. Proposed estimators have been compared with the sample median estimator when there is no matching from previous occasion and also compare with the ratio type estimator proposed by Singh et.al (2007) for second quartile. The behaviour of the proposed estimators are justified by empirical interpretation and validated by means of simulation study with the help of natural population. Keywords: Successive sampling, population median, bias, mean square error, auxiliary information.

EC-006 Forecasting Volatility of Selected Banks of Dhaka Stock Exchange (DSE), Bangladesh with GARCH (p, q) Types Models Hossain, S.1, M.A. Baten1 and F. B. Mukta1 1Department of Statistics, Shahjalal University of Science and Technology, Sylhet-3114, Bangladesh Email: [email protected], [email protected], [email protected]

Abstract: Excessive volatility in stock price fail to deliver the right signal on company’s real value although stock prices produce valuable information to safeguarding competence of capital markets. Stock price volatility is an indication of unreasonable market performance in Bangladesh. This study is to study the volatility and forecast future volatility of selected Banks of Dhaka Stock Exchange (DSE) by using different Generalized Autoregressive Conditional Heteroskedasticity GARCH(p,q) models. The volatility of DSE returns of four selected banks Brac bank, City Bank, Pubali bank and Eastern bank have been modelled for a period of 2008 to 2018. It is observed that the asymmetric volatility model E-GARCH and GJR-GARCH performs better to examine the existence of asymmetry in stock returns volatility of selected banks of DSE. There is a significant impact of positive shocks on increasing volatility For Brac and Eastern bank whereas for City bank negative shocks tends to produce higher volatility. Volatility is persistent for City, Pubali and Eastern bank which means that in upcoming days they could be affected by volatility. The range of 30 days forecasted volatility is found to be lower risk for both Brac bank and City bank comparing to Pubali bank and Eastern bank. Keywords: Volatility, Dhaka Stock Exchange, Selected Banks, GARCH, Forecast Submission for Student Paper Competition? Yes

EC-007 KNOWLEDGE AND AWARENESS ABOUT CERVICAL CANCER AND PREVALENCE OF HPV VACCINATION AMONG UNIVERSITY STUDENTS OF BANGLADESH Md. Abdur Rafi1, Kazi Abdul Mubin2, Md. Rezwanul Haque Choudhury3 Rajshahi Medical College, Rajshahi, Bangladesh, Email: [email protected] Department of Pharmacy, University of Rajshahi, Bangladesh, Email: [email protected] Rangpur Medical College, Rangpur, Bangladesh, Email: [email protected]

The incidence and mortality rate of cervical cancer in developing countries remain high despite being a preventable gynecological cancer by vaccination and early treatment due to poor awareness level. The present study aimed at assessing the knowledge of cervical cancer and identifying the HPV vaccination coverage among the young university students of Bangladesh. The cross sectional study was conducted among the students of University of Rajshahi during February, 2019 using a pretested self-administered structured questionnaire. χ² test was used to identify the predictors of knowledge level and binary logistic regression was used to identify the association between knowledge and taking HPV vaccine. Among 412 participants 60.4% was female and their mean age was 19.75 years. Most of the respondents believed

105

cervical cancer preventable but they had very little idea about the procedures of prevention or screening. Half of them were aware about HPV vaccine though most of them had no idea about the perfect age of vaccination. Almost one third of the respondents could not identify any risk factor and two third could not identify any symptoms of cervical cancer. HPV infection was most commonly identified risk factor followed by poor personal hygiene and multiple sexual partners. higher study years, older age, middle socioeconomic condition and positive family history of cervical cancer were significantly associated with better knowledge. Internet and social media was most popular source of information. About 15.3% female respondents had taken HPV vaccine. Lack of knowledge, opportunity and higher price were claimed as the commonest cause of not taking the vaccine. In regression analysis higher knowledge about cervical cancer was significantly associated with taking vaccine. Mass awareness programs should be launched to raise awareness about cervical cancer and HPV vaccination and the vaccine should be included national EPI schedule. Keywords: Cervical cancer, HPV vaccine, Awareness

EC-008 Association of Climatic Factor Variations with Recent Dengue Outbreak Identifies Challenges to Attain SDG. Susmita Sanyal MPH, North South University, Dhaka

Background: An outbreak of infectious diseases can potentially be correlated with climatic factors. The recent dengue epidemic can also be the result of climatic variation. Objective: To investigate the association between climatic-factor variations such as temperature, relative humidity and rainfall with dengue cases and deaths in the last 10 years to determine the correlation of climate change with dengue epidemic. Methods: Dengue cases and deaths were collected for 2009-2019 from DGHS. The climate data regarding temperature, relative humidity and rainfall in the period of 2009–2019 were collected from the Bangladesh Metrological Department. Linear regression model and correlation-coefficient analysis were used for statistical analyses. Results: Three different climatic-factors - temperature, humidity and rainfall were evaluated to be correlated with the recent high number of dengue cases and deaths. Among these three factors, temperature variation between 2019 and the last 4 years was observed to be positively correlated with the increased number of dengue-related cases and deaths. Conclusion: The outbreak of infectious diseases has been previously seen to be associated with temperature. The recent dengue epidemic turns out to be positively correlated as well like other infectious diseases like dengue or malaria. Keywords: Dengue outbreak, Relative humidity, Dengue epidemic, Climate change

EC-009 Big Data Analysis Using Sampling Techniques for the Finite Population Inference. Md. Ayub Ali1, Prof. Dr. Dulal Chandra Roy2 1Department of Statistics, University of Rajshahi, Bangladesh. 2 Department of Statistics, University of Rajshahi, Bangladesh. Corresponding E-Mail: [email protected]

In analyzing big data for finite population inference, it is critical to adjust for the selection bias in the big data. In this poster presentation, we use two methods of reducing the selection bias associated with the big data sample. We discuss how some of the sampling techniques can be applied in harnessing big data for finite population inference. By treating the selection bias in the big data sample as a missing data problem, we use two approaches of handling big data in survey sampling. The first method uses a version of inverse sampling by incorporating auxiliary information from external sources, and the second one borrows the idea of data integration by combining the big data sample with an independent probability sample. In this research work, an independent probability sample is used to estimate the parameters of the propensity score model for the big data sample. Two simulation studies show that the proposed methods are unbiased and have better coverage rates than their alternatives. In addition, these two methods are easy to implement in practice. Keywords: Inverse Sampling, Data integration, Non-Probability Sampling, Selection Bias

106

EC-010 The Performance of GARCH Modeling in Forecasting Groundwater Table Fluctuation of Northwest Bangladesh Md. Abdul Khalek, M. Sayedur Rahman and Md. Ayub Ali Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh. Corresponding E-mail: [email protected]

Nowadays, modelling and forecasting volatility of groundwater table have become significant interest to the practice of risk management. This paper used different volatility models; GARCH, GJR-GARCH, PGARCH, EGARCH and IGARCH to forecast groundwater table volatility of Northwest Bangladesh and find out the best model for forecast. This study uses weekly groundwater table data, collected from Bangladesh Water Development Board (BWDB) for the period January, 19991 to December, 2018. The period of January, 1991 to December, 2013 was used to build the model while remaining data were used to do out-sample forecasting and check the forecasting ability of the model. We find that the asymmetric model IGARCH following a normal error distribution yields the best forecasting performance. Our proposed model could be useful for forecasting groundwater table fluctuation of northwest Bangladesh, as well as leading to a better understanding of the groundwater table volatility, especially mitigating the problem of previous performance. Keyword: Groundwater Table; Volatility modelling; Volatility foresting; GARCH family model; Northwest Bangladesh.

EC-011 Logistic Regression based Feature Selection and Classification of Diabetes Disease using Machine Learning Paradigm Md. Maniruzzaman1, 2*, Md. Jahanur Rahman2, Benojir Ahammed1, Md. Menhazul Abedin1 1Statistics Discipline, Khulna University, Khulna-9208, Bangladesh; Email: 2Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh *Correspondence: Md. Maniruzzaman ([email protected])

Background and objectives: Diabetes is a chronic disease characterized by high blood sugar. It may cause many complicated disease like stroke, kidney failure, heart attack etc. There were about 422 million people affected by diabetes disease in worldwide in 2014. The figure will be reached 642 million in 2040. It is an alarming figure to us. The main objective of this study is to identify the most significant factors and proposed a machine learning (ML)-based system for predicting diabetes disease. Materials and methods: Logistic regression (LR) is used to identify the most significant factors for diabetes disease. Four types of ML-based system have adopted like naïve Bayes (NB), decision tree (DT), Adaboost (AB), and random forest (RF) to predict the diabetic disease. For the performance evaluation, five-fold (K5) cross-validation protocol is adopted and repeated this procedure in 20 trials. Performances of these classifiers are evaluated using accuracy (ACC), sensitivity (SE), positive predictive value (PPV), negative predictive value (NPV), and F-measure (FM), and area under the curve (AUC) computed from receiver operating curve (ROC). Results: National Health and Nutrition Examination Survey (NHANES) 2009-2012 diabetes dataset have been taken as a part of this study. The dataset consisted of 6561 respondents with 657 diabetics and 5904 controls. The overall accuracy of ML-based system is 90.62%. Moreover, the combination of LR and RF- based classifier gives 94.15% ACC, 99.54% SE, 94.29% PPV, 91.53% NPV, 96.84% FM and 0.94 AUC which are larger compared to others. Conclusion: The combination of LR and RF-based classifier performs better. This combination will be very helpful for detecting diabetic disease. Keywords: Diabetes, Machine learning, Naïve Bayes, Decision tree, Random forest, Adaboost, NHANES.

107

EC-012 Classification and Prediction of Parkinson Disease: A Machine Learning Approach Md. Menhazul Abedin1, Md.Maniruzzaman1, 2*, N. A. M. Faisal Ahmed1 Benojir Ahammed1, Mohammad Ali1 1Statistics Discipline, Khulna University, Khulna, Bangladesh 2Department of Statistics, University of Rajshahi, Rajshahi, Bangladesh *Correspondence: Md. Maniruzzaman ([email protected])

Parkinson's disease (PD) is a long-term degenerative disorder of the central nervous system that mainly affects the motor system. Nowadays classification and prediction of Parkinson’s diseases is a burning issue. We have used four machine learning techniques for Parkinson’s diseases classification and prediction. The data was extracted from UCI machine learning data repository. This research showed that support vector machine with polynomial kernel achieve accuracy, sensitivity, specificity, positive predictive value, negative predictive value 0.86, 0.85, 0.88, 0.87 and 0.85 respectively and they are the maximum values. This result makes the decision that support vector machine with polynomial kernel is the best classification and prediction method for Parkinson’s diseases. Keywords: Parkinson’s diseases, Classification, SVM, Accuracy, AUC.

EC-013 Prevalence and sex specific determinants of stunting among rural primary school children of Hooghly district, West Bengal, India Soma Pal1* and Kaushik Bose1 1Department of Anthropology, Vidyasagar University, Midnapore-721102, West Bengal, India. *Presenting author and corresponding author: Soma Pal, Ph. D. Research Scholar, UGC-NET (SRF), Department of Anthropology, Vidyasagar University, Midnapore - 721102, West Bengal, India. E-mail:[email protected]

Present study was undertaken to assess the prevalence of stunting and identify the sex specific risk factors of stunting among rural primary school children of Hooghly district, West Bengal, India. Height of 812 children (431 girls; 381 boys) was measured according to standard procedures. Date of birth was recorded from the school registers. Children were considered as stunted if HAZ was below minus two WHO Child Growth Standards 2007 (<-2SD). No sex difference was observed in mean height and HAZ. Age differences were significant in mean height for girls (F=117.597, p<0.001) as well as boys (F=104.184, p<0.001). Age differences in mean HAZ were not statistically significant for both sexes. The age combined prevalence of overall stunting among girls and boys were 20.2% and 19.4%, respectively. Age and sex combined prevalence of overall stunting was 19.8%. No significant sex differences in stunting were observed in each age group as well as age combined except at 6 years (χ2=6.517; df=1; p<0.05). Results of univariate Binary Logistic Regression analysis revealed that 9-10 years age group, Scheduled Caste (SC)/Scheduled Tribe (ST), 2 sibs, non-bricked house, no access of source of drinking water at home, tubewell as source of drinking water, manual category father occupation and working mother, ≤10th standards of father education had significantly high odds of stunting among studied girl pupils. Mid-Day-Meal consumption at school had significantly negative impact on stunting among girls. Stepwise multiple regression analysis revealed that household without access of drinking water source was important predictor for stunting among girls. The other risk factors of stunting among girls found were: working mother, older age (9-10 years), and ≤10th standards of father education. In boys, predictors of stunting were manual category father occupation and ≤10th standards of mother education remain in univariate binary logistic regression and stepwise multiple regression analysis. Prevalence of stunting was low as per public health perspectives. Household without access of drinking water source working mother, older age group, low father education were found to be predictors of stunting for girls. Manual father occupation and low mother education were observed as determinant of stunting among studied boys. Short Running Title: Stunting and associated factors among rural children KEY WORDS: Stunting; West Bengal; Primary school children, Rural.

108

EC-014 Inter Divisional Migration Flow of Bangladesh Md. Abdul Khalek, Md. Sabiruzzaman, Md. Mostafizur Rahman M. Sayedur Rahman and S. K. Bhattacharjee Department of Statistics, Rajshahi University, Bangladesh. Contact E-mail: [email protected]

This study uses 2001 and 2011 censuses data to measure the population flow by differencing present place of residence and place of birth with the help of CSPro version 6.0. The study shows that most of the migrants preferred to move in Dhaka. Majority of the migration flow was found from Khulna, Rajshahi, and Rangpur. The reverse flow from Dhaka is found to Sylhet division. The prediction of stability time of migration flows for each division is made using higher transition probability matrix. This study would help in achieving Sustainable Development Goals (SDGs) in relation to internal migration, urbanization, and combat climate change and its impact. Keywords: Migration, Migration flow, Inter division, Transition probability matrix.

EC-015 Association between exclusive breastfeeding and common childhood diseases in Bangladesh Mohammad Nayeem Hasan1, Najmun Nahar Nisha1 and MD Jamal Uddin1 1Department of Statistics, Shahjalal University of Science & Technology, Sylhet-3114, Bangladesh, [email protected] 1Department of Statistics, Shahjalal University of Science & Technology, Sylhet-3114, Bangladesh, [email protected] 1Department of Statistics, Shahjalal University of Science & Technology, Sylhet-3114, Bangladesh, [email protected]

Breastfeeding is a normal feeding method for infants and young children that ensures optimum growth and development. We aimed to assess the association between exclusive breastfeeding and common childhood diseases. We used the 2014 Bangladesh Demographic & Health Survey data and applied the Poisson model to analyze the data. Infants who were not exclusively breastfed (Risk Ratio (RR) 1.28, 95% CI 1.10–1.49, p = 0.0016) were found to be greater risk of having childhood diseases rather than who were exclusively breastfed. We found some other factors e.g., division, mother’s education, father’s occupation, mother’s body mass index, C-section and child’s sex were significantly associated with childhood diseases. In order to reduce child disease, still, proper breastfeeding is a vital issue in Bangladesh. Keywords: Exclusive breastfeeding, Childhood diseases, Poisson regression Submission for Student Paper? Yes

EC-016 Statistical Application for the Analysis of Traffic Congestion and Its Impact in a Hill City Utpal Kumar De1 and Gitumoni Rajbongshi2 1 Department of Economics, North Eastern Hill University, Shillong, 793022, Meghalaya, India Email: [email protected] 2 Department of Economics, North Eastern Hill University, Shillong, 793022, Meghalaya, India Email: [email protected]

Traffic congestion has been a severe problem of any growing urban area in the world that imposes drastic economic and social cost on the urban population. The livelihood activity and welfare are highly associated with the traffic flow pattern in the city. Over the years, with rising affluence, number of vehicles has increased disproportionately with road length and space. In East Khasi Hills district of Meghalaya, road length increased from 1696 KM in 2001 to merely 2050 KM in 2015, i.e., by 20.87 per cent. However, number of vehicles during the same period increased from 34996 to 137753, i.e., by 293.63 per cent. The problem is more acute in the hilly areas like Shillong due to scarcity of land space to enhance road width, hill topography and curvature along with difficulty and huge cost of construction. Steep hill slops and narrow space with rising vehicles in the peak hours block speedy movement of people, when there is an urgent need.

109

However, congestion varies across space and time of the day. It depends on topography of road, activity pattern and timing of vehicle use, locational pattern of institutions across length and breadth of the city. In this paper, an attempt has been made to statistically measure the level of congestion across some important road intersections on the national highway passing through Shillong city and find its major impact on the commuters. Analysis of congestion index and other indicators shows that there is traffic congestion across the Shillong city almost throughout the day excepting Sunday. The congestion becomes severe towards the intersection points surrounded by government offices and educational institutions especially during opening and closing hours of the offices. It has severe impact on the activity pattern, productivity and earning of the commuters with social and psychological consequences. For crossing a stretch of 250 meters around a crossing throughout the day, on an average a vehicle loses from 74 to 87 minutes of time. Of courses, the adverse effect of congestion varies across category of people in different occupation having diverse objectives of movements from home to regular destination involving varied opportunity cost. Keywords: Traffic Congestion, Impact of Congestion, Fuel Cost, Travel Time, Opportunity Cost, Shillong

EC-017 Growth – Inflation Relationship in Bangladesh: An Empirical Analysis for the Threshold Level of Inflation Shehan Uddin Khan Joint Director, Statistics Department, Bangladesh Bank Email: [email protected]

It is of significant economic substance to investigate the relationship of the inflation-growth nexus in Bangladesh, to recognize whether there is a relationship between the two macro variables. This can lead to the detection of the threshold level of inflation for Bangladesh. Accordingly, this research investigates the inflation-growth relationship in Bangladesh, using annual data from 1978 to 2018 fiscal year. Various econometric tests like augmented Dickey–Fuller test (ADF) tests, Breusch-Godfrey serial autocorrelation LM test etc. and Autoregressive Distributed Lag (ARDL) method are used for the growth equation. The results confirm significant structural break in per capita gross domestic product (GDP) growth rate when the inflation rate exceeds 6.5 per cent. However, This is an indication that any adverse effects of contemporaneous inflation are neutralized due to the significant positive effect of money supply (M2) and negative effects from the inflation lag of one and two years. Furthermore, the results are in favour of the view of maintaining inflation at low levels and thus, this study is important for policymakers, when implementing inflation targeting in Bangladesh in the future. Keywords: Economic growth, per capita GDP growth, inflation threshold.

EC-018 A Multilevel Analysis of Over-weight and Obesity among Non-Pregnant Women of Reproductive Age in Bangladesh: Evidence from BDHS data, 2014 Md Akhtarul Islam1, Hasin Raihan2 ,Md Nafiul Alam Khan 3 ,Abdul Mumin Bin Faruk 4 1Statistics Discipline, Khulna University,Email:[email protected] 2 Statistics Discipline, Khulna University, Email:[email protected] 3Statistics Discipline, Khulna University, Email:[email protected] 4Statistics Discipline, Khulna University, Email: [email protected]

Background: Over-weight and Obesity are a major public health challenge. It causes non-communicable diseases such as high blood pressure, type 2 diabetes, sleep apnoea, cancer, fatty liver diseases, pregnancy problems. Bangladesh is also facing a major crisis of overweight and obesity. In 1980, 7% of adults were overweight and obese in 2013, that climbed up to 17%. Objectives: We aim to explore the individual and community level factors associated with overweight and obesity among non-pregnant women of reproductive age. Compare to different geographical areas by multilevel analysis provide adaptive measures to policymakers and raise consciousness about overweight and obesity. Methods: Descriptive analysis and Multilevel models are used to find significant factors that are affecting over-weight and obesity using BDHS data,2014.

110

Results: At the individual level, results showed that (15-19) years of women from the poorest household had reduced odds of being overweight and obese. At the community level, women from urban areas and women who were less exposed to media have increased odds of being overweight and obese.78% of the total variance among the communities was explained by both community-level and individual factors. Conclusion: Women from the richest household should be targeted during policy-making at the community level. There lies the geographical effect of overweight and obesity at the community level and should be considered for nutrition policies. Keywords: BMI, multilevel, obesity, over-weight, Bangladesh Submission for Student Paper: Yes

EC-019 A Comparison of Logistic Regression and Linear Discriminant Analysis in Predicting of Female Students Attrition from School in Bangladesh Mohammad Nayeem Hasan1 1Department of Statistics, Shahjalal University of Science & Technology, Sylhet-3114, Bangladesh, [email protected]

This study aimed to compare the predictive accuracy and also the classification accuracy of linear discriminant analysis (LDA) and logistic regression analysis (LRA) using real data of school attrition. The overall classification accuracy for both models was determined by the classification accuracy rate. LRA and LDA classified 78.33% and 78.38% of girls respectively, in-school and out-of-school correctly. The AUROC curve for LRA was 80.63%, while it was 80.57% for the LDA. The LRA has sensitivity and specificity were 45.81% and 91.60%, respectively, and the LDA had a sensitivity and specificity of 46.81% and 91.01%, respectively. The overall classification rate for both was good. In comparison with the conventional LRA model, the LDA was better than LRA in the correct classification rate. In general, the results of this study showed that the LR model looks appropriate for prediction accuracy while LDA seems suitable to be used for classification techniques. Keywords: School dropout; Classification accuracy; Linear discriminant analysis; Logistic regression Submission for Student Paper? Yes

EC-020 Satisfaction with Work-Family Balance among Bankers: A Case-Study in Khulna City, Bangladesh Sheikh Rayhanul Islam, Mir Zubaer Ahmed Mst. Farzana Akter Reshmi and Md. Menhazul Abedin Khulna University, Bangladesh

Purpose: The paper seeks to examine the relationships between various work demands and resources and satisfaction with work-family balance in a sample of Bangladeshi bankers who were selected from Khulna city. Design/methodology/approach: A sample of 290 bank workers from both public and private banks in Khulna city participated in a comprehensive survey. Hierarchical multivariate regressions were mainly used to test the predicted relationships among the variables with the satisfaction with work family balance. Findings: From the work demand group perceived high organizational time expectations, psychological job demands were found to be positively related to bankers’ satisfaction with work-family balance and job insecurity were found to be negatively related at the same time. From work resources part social support at work and job control revealed positive relationships with satisfaction with work-family balance and the last factor which is work-to-home-interference showed negative relationships with the satisfaction with work- family balance. Research limitations/implications The study used bankers’ self-reports which may be problematic in drawing causal conclusions and another problem of our study was the low return rate of the delivered questionnaires to the bankers or to the bank managers. Originality/value: The majority of studies in work-family research look at either work-family conflict, or more recently, at work-family facilitation/enrichment having different professions’ employees, but little

111

research has been conducted on only bankers’ overall assessment of satisfaction with work-family balance. By investigating relationships between various work demands and resources in a sample of Bangladeshi bankers, the study contributes to the work-family literature by clarifying the relationship between work-to- family conflict and satisfaction with work-family balance.

EC-021 Landslide Susceptibility Zonation, Using Machine Learning Techniques, in a Part of the Hilly Areas of the Darjeeling District, West Bengal Diptarshi Mitra1 and Asim Ratan Ghosh2* 1Computer Science and Engineering Wing, Directorate of Distance Education, Annamalai University, Annamalainagar- 608002; current (residential) address: BB-35/5, Salt Lake City, Kolkata-700064; e-mail: [email protected] 2Department of Science and Technology and Biotechnology, Government of West Bengal, Bikash Bhawan (4th Floor), Salt Lake City, Kolkata-700091; e-mail: [email protected] *Corresponding Author

This study attempts Landslide Susceptibility Zonation of a part of the Kurseong Subdivision of the Darjeeling District in West Bengal, using Machine Learning techniques. Here, two models of Machine Learning, viz., Logistic Regression, implemented using Python programming, and Artificial Neural Network, implemented using Python and MATLAB programming, have been employed. And, four causative factors viz., land use/land cover, lithostratigraphy, structural features, and slope angle, have been considered, for susceptibility zonation. Also, data regarding the past landslide events in the study area, which occurred in or before 2015, have been used in this work. A part of the whole dataset pertaining to the causative factors and past landslide events, have been utilized for training and testing the models. While testing, the models exhibit high levels of accuracy. Then the models are applied to the whole dataset, and the output is recorded. For Logistic Regression model, and Artificial Neural Network model, implemented with Python, the probability is obtained as output. For Artificial Neural Network model, implemented with MATLAB, the output is adjusted to keep it within the range of 0 to 1, and this modified output is used as probability. Finally, the output is classified and converted into Landslide Susceptibility Zonation maps. The zones of highest susceptibility, in these Landslide Susceptibility Zonation maps, generated by the Logistic Regression and the Artificial Neural Network models, reasonably agree with the map of the previous landslide occurrences. Keywords: Landslide Susceptibility Zonation, Kurseong Subdivision, Logistic Regression, Artificial Neural Network, Python, MATLAB Submission for Student Paper/Poster Competition? No

EC-022 Cropping Pattern mapping of Suti - 1 Block using Machine Learning Technique Alamgir Shaikh1, Argha Ghosh2, Kajori Parial1 and Manoj K. Nanda2 1 West Bengal State Council of Science and Technology Salt Lake, Kolkata, West Bengal, India 2Department of Agricultural Meteorology and Physics, Bidhan Chandra Krishi Viswavidyalaya, West Bengal, India

With the global trend of increasing population, the demand for food is also expected to increase. The traditional way of increasing the crop production simply by adding up new lands for agricultural activities is no more feasible owing to the competition for land to facilitate other human activities. In this scenario, the alternative is to increase the produce from the same amount of or possibly from less amount of land through replacing monoculture farms with diversified farms. To increase the production through diversified agriculture, a proper crop planning is required. To carry out a proper crop planning, it is essential to know the existing cropping pattern. In this present study the cropping system of Suti-1 block of Murshidabad district has been assessed through multi dated satellite images. Spectral profiles of different cropping systems were generated to understand the conditions of the crops in this study area. Normalized Difference Vegetation Index (NDVI) was calculated for each and every cropping system to study the developmental cycle of the crops. Cropping systems could be successfully differentiated by studying the NDVI profile of the systems. The findings of the present study gave an overview of the cropping scenario of the study area. Keywords: Cropping pattern, Random forest, NDVI, Sustainable agriculture. Submission for Student Paper/Poster Competition? No

112

EC-024 Relationships among Statistics Anxiety, Depression and Academic Performance Kawsar, L. A.1* , Zahan, F. N.2 , and Islam, M. A.3 1* Department of Statistics, Shahjalal University of Science and Technology, Sylhet, Bangladesh, Email: [email protected] 2 Department of Statistics, Shahjalal University of Science and Technology, Sylhet, Bangladesh, Email: [email protected] 3 Department of Statistics, Shahjalal University of Science and Technology, Sylhet, Bangladesh, Email: [email protected]

Statistics anxiety is one of the academic stresses faced by students. It has been identified in students with an adverse effect on their learning and performance. Statistics anxiety can adversely affect students’ overall psychological condition and hence cause depression in them. Various measures have been developed in order to identify and assess statistics anxiety among students. Particularly no studies have focused on its relation with depression. Considerable amount of studies related to statistics anxiety has been done in English-speaking populations. However, no study has been done regarding this topic in Bangladesh. Therefore, this study aims to investigate the relationship between statistics anxiety, performance and depression among under-graduate students of Shahjalal University of Science and Technology, Sylhet, Bangladesh. Data was collected from a total of 411 undergraduate students who had enrolled Statistics course as a non-major course in their running semester during the months of January, 2019 to Jun, 2019 of SUST using statistics anxiety rating scale (STARS) and PHQ-9 (measure for depressive symptoms) questionnaires. It is found that 84.42% students were suffering from some sort of statistics anxiety. It is also observed that 85.4% students were dealing with some sort of depression. Among them 46% students were dealing with major depression. Statistics anxiety has been found to be significantly effecting students’ performance and depression level (p-value <.001). Students suffering from high level of statistics anxiety were found to be suffering from high level of depression and attaining poor marks in academic exams. Statistics anxiety is found to affect students’ performance and increase the level of depression in them. Findings of the present study would be helpful for the students, teachers, even policy makers to design student friendly curriculum to reduce the statistics anxiety among students in future. Keywords: Statistics anxiety, Depression, STARS, PHQ-9. Submission for Student Paper/Poster Competition? Yes

EC-025 Awareness of Diarrhoea on pregnant Women in Bagerhat district Md. Abu Borhan1 and Md. Ayub Ali2 1 MPSc in Applied Statistics, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh 2 Professor, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh

Background: The World Health Organization defined that diarrhoea is the passage three or more lose or liquid stools per day (or more frequently than is normal for the individual). The analysis by the World Health Organization shows that diarrhoeal disease is the second leading cause of death in children under five years old. Globally, there are nearly 1.7 billion cases of childhood diarrhoeal disease every year. Diarrhoea is a leading cause of malnutrition in children under five years old. Awareness of pregnant women on diarrhoea reduce mortality and morbidity rate of pregnant women and postnatal children. Objectives: The purpose of the present study was to find out the awareness pattern on diarrhoea in pregnant women of Bagerhat district. Materials and Method: A sample of 29 pregnant women (PW) from a total of listed 111 women from three upazilas of Bagerhat district was considered for assessing the awareness of diarrhoea. Those three upazilas were taken at random first from the nine upazilas of Bagerhat district. Data on different variables were collected directly from the selected women through a prescribed questionnaire. Descriptive statistics e.g., maximum, minimum, mean, standard deviation, skewnwss, kurtosis, etc of the variables together with their standard error of estimates were considered for analyzing sample characteristics of the study. The relationship between two nominal variables is assessed by cross tabulation with test statistics Phi, Cramer's V and coefficient of contingency.

113

Results: About 62.1% pregnant women in Bagerhat district have no awareness about diarrhoea. Also, they do not know about the causes of diarrhoea, knowledge management of pregnant women on diarrhoea prevention The test statistics Cramer's V, coefficient of contingency and Phi Correlation coefficient divulged that the relationships between knowledge management and knowledge about diarrhoea, reasons for diarrhoea and knowledge about diarrhoea, knowledge management and sign-symptoms of the disease, and reasons for diarrhoea and sign-symptoms of the disease were all statistically insignificant at 5% level of signficance which is a contradiction to the fact that the pregnant women over Bagerhat district have pure knowledge about diarrhoea and diarrhoeal diseases. Conclusion: The awareness on diarrhoea of pregnant women in Bagerhat district should be more authentic so that no pregnant women as well as their child wouldn't be affected with diarrhoea and diarrhoeal diseases. Recommendation: Government should ensure proper awareness on diarrhoea of the pregnant women in Bagerhat district and may be in the other district also. Keywords: Awareness, Diarrhoea, Pregnant women, Phi Correlation and Cramer's V

EC-026 Statistical analysis on mid upper arm circumference of school going children (6-10 years) in Rajshahi district, Bangladesh

Md. Ismail Hossain*, ASMA Mamun, Md. Nurul Islam Md. Ripter Hossain and Md. Golam Hossain Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh *Md. Ismail Hossain, Ph.D. Fellow, Department of Statistics, University of Rajshahi, Bangladesh Email: [email protected]

Nutrition, genes and environment are the three major factors that impact a child’s development. Nutritional status of children is an important indicator of their health status of a particular population. According to the World Health Organization (WHO), malnutrition is the gravest single threat to global public health. Globally, it contributes to 45% of deaths of children aged under five children, and most of the death occurs in developing countries. Study on nutritional status of children is important for determining risk factors and current prevalence of malnutrition in a particular country to take initiative for remove or reduce malnutrition among children. Many studies have been done with under five children nutritional status in worldwide including Bangladesh, but few studies are available on nutritional status of school going children aged 6-10. To the best of our knowledge the study on nutritional status of school going children aged 6-10 years in Rajshahi district is poorly documented. The objectives of the present study was to know the nutritional status and to identify the risk factors among school going children aged 6-10 years in Rajshahi district, Bangladesh. Sample was collected from the different selected schools of urban and rural area in Rajshahi district, Bangladesh using multistage stratified sampling. A total number 1200 school going children aged 6-10 years were selected as a sample for this study. Data were collected from selected students and their parents using a standard questionnaire from May to October 2015. The nutritional status of children was measured by mid upper arm circumference (MUAC). The cut-off points of the anthropometric measurement were used to classify the nutritional status such as normal and under nutrition. Frequency distribution, Student’s t-test, ANOVA, multiple linear regression, Chi-square test, logistic regression and Stepwise regression analysis were utilized to serve the purpose of the objectives for the present study. The mean age of the student was 7.77±1.40 year with range 6 to 10 years. The mean MUAC of school going children was 18.20±2.85 cm with an increasing tendency was observed with increasing their age. It was found that student’s HIP, student’s BMI, student’s age, father’s age, head cercumference (HC) of student, and family monthly income were the most important predictors of children MUAC. Chi-square test demonstrated that more than 80% of our selected factors were associated with nutritional status of school going children aged 6- 10 years in Rajshahi, Bangladesh. In this study, we found some modifiable risk factors of malnutrition among school going children aged 6-10 years in Rajshahi district, Bangladesh. Consequently, malnutrition can be considered as a major health problems of Bangladeshi school going children aged 6-10 years and require attention. KEYWORDS: Malnutrition, Multiple regressions, School going children, Rajshahi, Bangladesh

114

EC-027 Factors Associated with Availability of Selected Essential Medicines for Diabetes and Cardiovascular Diseases in Health Facilities of Bangladesh Shariful Hakim1, Md. Ashiqul Haque2, Muhammad Abdul Baker Chowdhury3 and Md Jamal Uddin4 1 Lecturer (Statistics), Chander Hat Degree College, Nilphamari, Bangladesh. Email: [email protected] 2 Graduate Student, University of New Brunswick, Fredericton, NB, Canada. Email: [email protected] 3 Department of Biostatistics, Florida International University, FL, USA. Email: [email protected] 4 Department of Statistics, Shahjalal University of Science & Technology, Sylhet, Bangladesh. Email: [email protected]

Background: The burden of diabetes and cardiovascular diseases (CVD) continues to increase in developing countries including Bangladesh. One of the main factors responsible for this burden is the availability and accessibility of essential medicines of the diseases. However, there is not much research done on the availability of essential medicines for these diseases in Bangladesh. Objective: To assess the availability of essential medicines for diabetes and CVD and to identify health facility characteristics associated with availability of those medicines. Methods: The 2014 Bangladesh Health Facility Survey (BHFS) data was used in this study. The 2014 BHFS is the third survey of its kind in Bangladesh and was designed to provide national-level information for the various types of health facilities. Based on our inclusion criteria (e.g., hospitals ≥20 beds), information from 275 facilities for diabetes and 250 facilities for CVD was used in the analyses. The outcome variables, essential medicines for diabetes was calculated as a count score of ranging from 0 to 4 and essential medicine for CVD was also calculated as count score of ranging from 0 to 5, by which higher scores indicated the greater availability of essential medicines. The possible independent variables were place of residence, division of residence, managing authority, inpatient only, guidelines for diagnosis and management), trained staff. Data were analyzed using Poisson regression model. Results: Of the selected facilities for diabetes, 238 (86.6%) facilities had no essential medicine for diabetes (EM-Diabetes) and of the selected 250 facilities for CVD, 182 (72.8%) facilities had no essential medicine for CVD (EM-CVD). Only 9.8 percent of facilities that offer diabetes services had metformin available on the day of the survey. Among facilities that offer cardiovascular disease services, 21.6 percent had beta blockers available on the day of the survey. The risk for availability of EM-Diabetes was 7.54 higher in those facilities which situated in urban areas compared to those situated in rural areas (Relative Risk (RR): 7.54; Confidence Interval (CI): 3.35-17.12). The risk for availability of EM-Diabetes was 1.57 higher in facilities that had guidelines for the diagnosis and management compared to facilities that had no guidelines for the diagnosis and management of diabetes. (RR: 1.57; CI: 0.99-2.48). The risk for availability of EM- CVD was 2.83 higher in those facilities situated in Sylhet division compared to those situated in Barisal. (RR: 2.83; CI: 1.19-6.75). The risk for availability of EM-CVD was 1.23 higher in facilities that provide inpatient care service compared to facilities that do not provide inpatient care service (RR: 1.23; CI: 0.85- 1.79). Conclusion: There are vast disparities between rural and urban facilities in availability of these essential medicines. Patients with diabetes, and cardiovascular disease need a reliable supply of essential medicines. An allegiance by governments to fulfill the needs of their citizens who suffer from diabetes and CVD is pressingly needed. Keywords: Diabetes, Cardiovascular diseases, Essential medicine, BHFS

EC-028 Two level logistic regression analysis of factors influencing dual form of malnutrition in mother-child pairs: A household study in Bangladesh Md. Akhtaruzzaman Limon*, ASMA Mamun, Kumkum Yeasmin Md. Moidul Islam and Md. Golam Hossain Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh * Presenter, E-mail: [email protected]

Bangladesh is undergoing a nutrition transition associated with rapid social and economic transitions giving rise to the double burden of the malnutrition phenomenon. It is essential to investigate the household study of malnutrition among mothers and under-five children pair. The objective of this study was to determine the prevalence and risk factors of malnutrition among mother-child pairs at the same household in Bangladesh.

115

Secondary data from the BDHS – 2014 was used in this study. The sample population of this study consisted of 7,368 married, currently non-pregnant Bangladeshi women with their under-five child. Descriptive statistics, Chi-square (χ2) tests and two-level binary logistic regression model was used in this study. The prevalence of underweight mother-child pair was 22.0% and the prevalence of overweight mother and underweight child were near to 10%. It was found that only less than 20 percent (19.6%) mother and child pair was found to be normal weight (healthy). The two-level binary logistic model showed that division, type of residence, parents’ education, household wealth index, mothers’ age and children birth weight were found to have risk factors of under nutrition among mother-child pair. Our selected model proved the risk factors of under nutrition among mother-child pairs in Bangladesh. These factors can be considered for reducing the number of malnutrition among mother-child pairs in Bangladesh. Keywords: Nutritional status, Multilevel logistic regression, Mother-child pair, Bangladesh

EC-029 Factors Affecting Birth Weight of a Newborn – A Community Based Study in Gopalganj, Bangladesh Md. Murad Hossain1, Syeda Nazmun Nahar2 and Md. Minhajul Islam3 1 Department of Statistics, Faculty of Science, Bangabandhu Sheikh Mujibur Rahman Science and Technology University, Gopalganj-8100, Bangladesh. Email:[email protected] 2,3 Department of Statistics, M.Sc. student Bangabandhu Sheikh Mujibur Rahman Science and Technology University,Gopalganj-8100, Bangladesh. Email:[email protected],[email protected]

Birth weight is one of the indicator of newborn health. Most of the people do not have proper knowledge about birth weight. In this paper we tried to know which factors are associated with the birth weight status of new born. This was a survey based study from different place such as Government hospital, private hospital and different area. This information was collected from the mother by pre tested questionnaire. We collect 150 new born weight at 0-6 month and all birth weight were measured within 24 hours of delivery. The variables age of the pregnant women, pregnancy time, day time rest during pregnancy, antenatal care have been considered important factor for new born birth weight. The dependent variable was chosen birth weight. Bivariate analysis is used to assess the association between birth weight and list of independent variables. Binary logistic model was applied to detect the impact of factor the independent variable of the Pearson chi-square test. All statistical analysis was performed using SPSS version 20 software. A p-value <.05 was considered significant. Key Word: Birth Weight, pregnancy time, rest during pregnancy, antenatal care.

EC-030 Statistical Analysis of Various Optimal Latin Hypercube Designs A R M Jalal Uddin Jamali*,1, Md. Asadul Alom2, Abdul Aziz3 1Department of Mathematics, Khulna University of Engineering & Technology, BD 2Daud Public College, Jessore, BD 3Department of Computer Science and Engineering, Khulna University of Engineering & Technology, BD *Corresponding Author: Email: [email protected]

Abstract. Among several Design of Experiments (DoEs,) Latin Hypercude Design (LHD) is one of the most frequently used method in the field of physical experiments and in the field of computer simulations to find out the behaviour of response surface of surrogate model with respect to design points. A good experimental design should provide three important characteristics namely (i) Non-collapsing, (ii) space- filling (iii) Orthogonal property. Though inherently LHD preserves Non-collapsing property but randomly generated LHDs have poor space-filling in term of minimum pair-wise distance. In order to cover these other two characteristics of LHD, researchers are frequently look for finding optimal LHD in the sense of space- filling and orthogonal properties. In the literature several type of Optimal LHDs are available such as maximin LHD, Orthogonal LHD, Uniform LHD etc. In this article intensive statistical analysis has been carried out upon numerical instances to view the insight of each optimal LHD. It seems to us that in the sense of space-filling Maximin LHD is the best but for both space-filling and orthogonal properties multi- objective (maximin with approximate orthogonal) based optimal LHD is the best. Keywords: Experimental Design, Space-filling, Non-collapsing, Orthogonal, Euclidean Distance. Submission for student Paper/Poster? No

116

EC-031 Association between Campylobacter jejuni/coli infection and childhood stunting in children aged 0-24 months in low-resource settings Md Ahshanul Haque1, Md. Ashraful Alam1, Mustafa Mahfuz1, Tahmeed Ahmed1 1Nutrition and Clinical Services Division, icddr,b

Background: Campylobacter infection is associated with malnutrition among children in low-resource settings. It remains unclear whether that association is specific to Campylobacter jejuni/coli. The aim of this study is to assess the association between Campylobacter jejuni/coli infection and childhood stunting in children aged 0-24 months in low-resource settings. Methods: Data of 1715 children from 0-24 months were derived from the MAL-ED birth cohort study in Bangladesh, Brazil, India, Nepal, Pakistan, Peru, South Africa, and Tanzania. Campylobacter jejuni/coli were detected in stool samples using the molecular test called TaqMan Array Cards (TAC). Descriptive statistics served to illustrate the socio-demographic variables. Associations between Campylobacter jejuni/coli infection and childhood stunting were estimated using generalized estimating equations to fit regression models with binomial family after adjusting for seasonality, sex, water/sanitation, assets, maternal education, and income index; enrollment length-for-age; maternal height; poultry in house, and site for overall estimate and age in month as time variable. Results: The incidence rate per 100 child-months of infections with Campylobacter jejuni/coli during 1-24 month follow up was 17.7%. Overall, the prevalence of stunting among the Campylobacter jejuni/coli infected participants was 34% and 26% among those who were not infected. Children infected with Campylobacter jejuni/coli were 10% [OR: 1.10 (95%CI: 1.03, 1.17), p=0.003] more likely to be stunted compare to those who were not infected. In Bangladesh and Tanzania, such likelihood was higher, at 37% [OR: 1.37 (95% CI: 1.20, 1.58), p=0.000] and 42% [OR: 1.42 (95% CI: 1.25, 1.62), p=0.000] respectively. Conclusion: Campylobacter jejuni/coli infection was associated with poor growth; such infection is higher in Bangladesh and Tanzania compare to other sites. Effective interventions, policies need to be designed to minimise these risk factors and enhance or improve children growth and achieve better developmental outcomes. Keywords: Enteropathogens, Campylobacter jejuni/coli, Children, Malnutrition, Quantitative PCR

EC-032 Study on knowledge about tuberculosis of female sex workers in Rajshahi city, Bangladesh Md. Imran Ali1*, Anamika Saha2, Md. Masud Rana2, Md. Rafiqul Islam2 Abu Sayed Md. Al Mamun1, Md. Golam Hossain1 1Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh 2Department of Population Science and Human Resource Development, University of Rajshahi, Rajshahi- 6205, Bangladesh * Email: imrannazir000125@gmail .com,

Female sex workers (FSWs) are one the most important vulnerable high risk group for getting and spreading TB. The study on lack of general knowledge of TB among FSWs was poorly documented in Bangladesh. The object of this study was to identify the risk factors for lacking knowledge on TB among FSWs in Rajshahi city, Bangladesh. The sample consisted of 225 FSWs who traded sex in Rajshahi city, Bangladesh. Data were collected from July’2015 to December’2016. This study was based on the complete good clinical procedures. Chi-square test was utilized to find the association between knowledge on TB and selected socio-demographic factors. Out of 225 FSWs, 43.1%, 34.7% and 22.2% came from urban, rural and slum areas respectively. More than 41% FSWs perceived that TB is a non-communicable disease. A large number of (76.4%) FSWs did not knew how TB spreads, and most of FSWs (90%) had no knowledge on latent TB. χ2-test demonstrated that FSWs’ education, monthly income and sex trading place were associated with their knowledge on TB. It was found that comparatively more educated, living with rich family, currently unmarried, traded sex at residence, aged≥26 years FSWs were more knowledgeable than their counterparts. This study strongly recommends that advocacy, communication and social mobilization program is urgently needed for FSWs. Keywords: Tuberculosis, female sex workers, χ2-test and Bangladesh. Submission for Student Poster Competition? Yes

117

EC-033 Double burden of anemia among mothers and their under-five children in Bangladesh Mst. Shela Khatun, Md. Abdul Wadood1, Md. Rezaul Karim1 Abu Sayed Md. Al Mamun1, and Md. Golam Hossain1 1Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh

The household study on presence of anemia is poorly documented. This study was conducted to determine dual burden of anemia among the mothers and their under-five children pairs in Bangladesh and investigate the associated risk factors. Data was extracted from Bangladesh Demographic and Health Survey 2011. A total of 1,281 pairs of mothers of reproductive age (15-49 years) and their under-five children (6-59 months) were considered as samples. Frequency distribution and chi-square test and binary logistic regression model were used in this study. The dual burden of anemia among mothers and their under-five children was found to be 45.8%. Currently breastfeeding status, religion, household wealth index, number of total ever born children, children’s age and mothers’ nutritional status were the most important predicators (95% confidence interval) for anemia. Situation of dual burden of anemia was found to be alarming in Bangladesh. Identified risk factors should be tackled to reduce prevalence of anemia for minimizing maternal and child morbidity and mortality which are the two important components of the SDG. Antenatal and postnatal care and food supplementation for pregnant and lactating mothers and their children especially in underprivileged regions and population groups should be actively considered. Keywords: Anemia, Mothers, Children, Public Health, Bangladesh Submission for Student Poster Competition? Yes

EC-034 Factors influencing exclusive breastfeeding practice in Bangladesh: Multiple logistic regression analysis Shobnom AffifaTanny1, ASMA Mamun1, Md. Sabiruzzaman1 and Md. Golam Hossain1 1Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh. E-mail: [email protected], [email protected], [email protected] and [email protected]

Breastfeeding protest against diarrhea and common child illness such as pneumonia and may also have long term health benefits for mother and child such as reducing the risk of overweight and obesity in childhood and adolescence. WHO estimates that around 220,000 children could be save every year with exclusive breastfeeding. Thereafter, to meet their evolving nutritional requirements, infants should receive nutritionally adequate and safe complementary foods, while continuing to breastfeed for up to two years or beyond. The objective of our study was to determine the prevalence and associated factors of exclusive breastfeeding for first six months of an infant’s life in Bangladesh. Data used in this study that was extracted from Bangladesh Demographic and Health Survey (BDHS-2014) 2014. A total of 3541 non-pregnant married women were considered as the sample in the present study. Chi-square (χ2-test) and multiple logistic regression models were utilized in this study. It was found that the prevalence of exclusive breastfeeding among Bangladeshi non-pregnant mothers was 35.9% (urban 35.17% and rural 24.56%). The multiple binary logistic regression models demonstrated that the non-pregnant mothers who lived in Sylhet division were more likely to provide EBF to their infants than those in Dhaka (AOR=0.669; 95% CI: 0.524-0.855, p<0.05) and Rajshahi divisions (AOR=0.745; 95% CI: 0.571 - 0.855, p<0.05). It was noted that uneducated mothers were more likely to provide EBF to their infants than higher educated mothers (AOR=1.427, 95% CI: 1.111-1.949; p<0.05). The mothers who accessed mass media were more likely to provide EBF (AOR=0.1.169; 95% CI: 1.015-1.347, p<0.05). It was observed that housewife mothers were more likely to give EBF to their infants than the mothers who worked outside of home (AOR=0.763; 95% CI: 0.652-0.893, p<0.05). Finally, step-wise logistic regression model demonstrated that living location (division), mothers’ education level, mass media access, total number of ever born children, baby PNC, mothers’ occupation, husbands’ occupation, child current age were most important predictors for proving EBF among Bangladeshi mothers. It was observed that near to 65% mothers did not provide EBF to your infants. This study suggests that government and non-government health authorities should take initiative steps for making awareness about the benefits of exclusive breastfeeding among Bangladeshi non-pregnant mothers. Keywords: Exclusive breastfeeding, Bangladesh, Non-pregnant mothers, Logistic regression Submission for Student Paper Competition? No

118

EC-035 Prevalence and associated factors of stunting among school going children (6-13 years): Survey in rural area of Rajshahi district, Bangladesh Md. Reazul Karim, Md. Ripter Hossai1, Md. Nurul Islam Abu Sayed Md. Al Mamun, and Md. Golam Hossain Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh

Stunting is one of the most serious public health problems among school going children in Bangladesh. It is increase the risk to survival, development, school performance and productivity in adult life, which constitutes a significant obstacle to achieving better child health outcomes. This study aimed to determine the prevalence and risk factors of stunting among school going children aged 6–13 years in rural area of Rajshahi district, Bangladesh. It was a cross-sectional study with a sample size of 500 school going children aged 6-13 years in high Barind tract area in Rajshahi district of Bangladesh. Data were collected from the field survey conducted from January to March, 2019. Multistage cluster sampling was used for selecting our sample from the population. A standard questionnaire was prepared for collecting information regarding child nutritional status and others variables. Chi-square (χ2)-test and binary logistic regression were used to find the effects of socio-demographic and physical activities factors on children stunting. Statistical significance was measured using adjusted odds ratio (AOR) at 95% CI and P value less than 0.05. This study revealed that the prevalence of stunting was 28.4%. Younger mother’s (age ≤30 years) children were more like to suffer with stunting than others (AOR= 0.571; 95%, CI: 0.338-0.964, p< 0.05). Similarly found that mothers’ those weight <45 kg were more likely to have stunting children than mother those weight were more 45 kg (AOR= 0.518; 95%, CI: 0.274-0.981, p<0.05). Findings suggest that maternal nutrition, nutritional intervention in school for rural school going children should ensure for reducing stunting. Keywords: School going children aged 6-13, stunting, socio-demographic, logistic regression Submission for Student Paper Competition? Yes

EC-036 Factors associated with Tuberculosis: a case control study in Rajshahi city, Bangladesh Md. Masud Rana1, Md. Rafiqul Islam1, Md. Abdul. Wadood2 Abu Sayed Md. Al Mamun3and Md. Golam Hossain3 1Department of Population Science and Human Resource Development, University of Rajshahi, Rajshahi-6205, Bangladesh, E-mail: [email protected], [email protected] 2Rajshahi University Medical Center, University of Rajshahi, Rajshahi-6205, Bangladesh, Email: [email protected] 3Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh, E-mail: [email protected], [email protected]

Tuberculosis (TB) is a major public health problem in Bangladesh since long. Though various measures have been taken, morbidity and mortality due to TB are still high for some socio economic conditions prevailing in this country. The purpose of this study was to investigate the associated factors of TB. It was a case- control study. Study data were collected from January’2017 to July’2017. The sample consisted of 268 TB patients (case) and 268 close contacts (who did not get TB) of TB patients (control). Most of the non TB closes contacts persons were TB patient’s siblings living in same location in Rajshahi city, Bangladesh. Chi- square test and binary logistic regression model were used to determine the risk factors of TB. Out of the total samples, approximately 76% were above 26 years of age from case and control groups. 58.6% and 41.4% were male and female in case group, 65.3% and 34.7% were male and female in control group respectively. Through marital status, 12.7% and 87.3% unmarried and ever married in case group, 11.6% and 88.4% unmarried and ever married in control group respectively. More than 42% TB patients were under nourished while only 10.1% their close contacts were under nourished. Logistic regression provided that under nourished individuals were 8.253 times more risk for getting TB [(AOR=8.253; 95% CI: 4.251- 16.019, p< 0.01] than over nourished. It was noted that uneducated were more likely for getting TB than secondary and higher educated adults [AOR=1.864; 95% CI: 1.148-3.026, p< 0.05] and primary educated adults were 1.532 times higher for getting TB than Secondary and higher educated adults [AOR=1.532; 95% CI: 1.051-2.234, p< 0.05]. This study strongly recommended advocacy communication and social mobilization and social safety net program for nutritional life led and education level will be increased. Keywords: Tuberculosis, Rajshahi city, Chi square test and Binary logistic regression Submission for Student Paper Competition? No

119

EC-037 Risk factors of bipolar disorder among married women: a household study in Rajshahi city, Bangladesh Md. Abdul Wadood1, Md. Rezaul Karim2, Md. Masud Rana3 Abu Sayed Md. Al Mamun4, and Md. Golam Hossain4 1Medical Centre, University of Rajshahi, Rajshahi- 6205, Bangladesh. 2Department of Biochemistry and Molecular Biology, University of Rajshahi, Rajshahi- 6205, Bangladesh. 3Deparment of Population Science and Human Resource Development, University of Rajshahi,, Bangladesh. 4Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi- 6205, Bangladesh.

Bipolar disorder (BPD), a major mental illness, affects personal and social functioning and creates huge economic burden. It is poorly documented in Bangladesh. This study was designed to determine its prevalence and investigate the associated factors among married women in Rajshahi city, Bangladesh. It was a household cross-sectional study. Data were collected from 9 muhallas in Rajshahi city by multistage stratified systematic sampling method. The sample size was 279 married women. Bipolar spectrum diagnostic scale was used for screening BPD. Frequency distribution, chi-square test and binary logistic regression model were used to determine the prevalence, and investigate the associated factors and effects of the factors on BPD respectively. The prevalence of BPD was found to be 2.5%, and mental disease (AOR=0.120, 95 % CI: 0.020-0.716; p<0.05), poor relation with husband (AOR=0.085, 95 % CI: 0.015- 0.483; p<0.01), poverty (AOR=1.600, 95 % CI: 2.086-122.709; p<0.01), high education level (AOR=0.177, 95 % CI: 0.037-0.843; p<0.05), not immediately treated if sick (AOR=2.941, 95 % CI: 1.259-6.871; p<0.05) and death of beloved one/s (AOR=2.768, 95 % CI: 1.130-6.777; p<0.05) were the main predictors of BPD. This study laid down the ground-stone for further research on different aspects and issues of BPD in Bangladesh, such as, familial environment and culture, comorbidities, treatment options, treatment outcomes, biochemical picture, environmental factors, etc. This study also suggests that health professionals should focus on comorbidities and family matters while giving treatment to the BPD patients. Keywords: Bipolar disorder, Married women, Risk factors, Rajshahi, Bangladesh Submission for Student Paper Competition? Yes

EC-038 Factors associated with lack of general knowledge on HIV/AIDS among female sex workers of Rajshahi City in Bangladesh Astami Devnath1, Md. Jahangir Alam1, Sheikh Moin Uddin2 Abu Sayed Md. Al Mamun1 and Md. Golam Hossain1 1Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh 2Faridpur TB and Leprosy Control Project, Damien Foundation, Faridpur, Bangladesh

In worldwide, female sex workers are considered as a vulnerable group for possessing HIV/AIDS. Female sex workers (FSWs) and their clients are the major sources of spread of HIV/AIDS. In Rajshahi there is no brothel, female sex workers trading sex with their clients at different places. It was needed to investigate the knowledge on HIV/AID among female sex workers in Rajshahi city, Bangladesh. The aim of this study was to find the risk factors of lack of general knowledge on HIV/AID among female sex workers in Rajshahi city, Bangladesh. Data were collected from female sex workers who were trading sex in Rajshahi city, Bangladesh. Female sex workers were contacted with the cooperation of the different organization who was working for the welfare of female sex workers in Rajshahi city, Bangladesh. This study was designed according to the principles of good clinical practice. Data was collected from July to September, 2018. Frequency distribution, Chi-square test and logistic regression models were used in this study. A total number of 425 female sex workers were selected randomly. The age range of female sex workers was 16-41 years (average age, 24.52±6.26 years). It was found that near to 50% female sex workers were primary educated and 33.6% were secondary educated. More than 36% and 30% came from rural and urban areas reactively, while 33.2% female sex workers were living in slum environment. It was found that more than half of the female sex workers (56.9%) were in age group 21-29 years, while 22.6% were young (age≤20 years) and 20.5% comparatively old (age≥30 years). Our data revealed that 88% FSWs are married. More than 88% of FSWs reported to have practicing unprotected sex because of client’s insistence. 89.5% female sex workers did not ask their new clients about their HIV test status. Further analysis showed a significantly

120

higher rate of protected sex among unmarried FSWs (p<0.01). Approximately 32% of FSWs did not undergo voluntary counseling and testing (VCT) for HIV. Logistic regression analysis showed that FSWs who traded sex at residence, already tested for STI, were in the age group 21-29 years, were secondary and higher educated, came from rich family, were interested to use condom and asked new clients about HIV testing. Health authorities of Bangladesh should focus their HIV/AIDS prevention efforts on illiterate, married and poor FSWs. Consequently, since FSWs are considered as a vulnerable group for spreading HIV/AIDS, authority should take steps to promote HIV testing for FSWs in this country. Keywords: HIV/AIDS, chi-square test, logistic regression, Rajshahi City and Bangladesh Submission for Student Poster Competition? Yes

EC-039 Factors influencing unhygienic place of delivery among Bangladeshi mothers: Data from Bangladesh Demographic and Health Survey, 2014 1Md. Abu Sayem, 2Md. Israfil Hossain, 2Md. Nurul Islam, 2 Abu Sayed Md. Al Mamun Mamun, 2Md. Golam Hossain 1EXIM Bank Agricultural University Bangladesh, Chapainawabganj-6300, Bangladesh 2Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh

Home delivery in unhygienic place is common among Bangladeshi women. Delivery related complication is one of the leading causes of maternal and neonatal mortality and morbidity. The objective of this study was to investigate the place of delivery among Bangladeshi mothers. The data for this study were extracted from Bangladesh Demographic and Health Survey (BDHS-2014). Data for 4273 Bangladeshi mothers having at least one live birth were included for analysis.The chi-square test and multinomial logistic regression model were used in this study. The prevalence of home delivered mothers without proper nursing in Bangladesh was 59.3% (67.7% rural and 41.9% urban). The multinomial logistic regression analysis showed that mothers who lived in rural area were more likely to perform home delivery (p<0.05) than urban mothers. Mothers lived in Sylhet divisions were more likely to have home delivery than in Khulna (p<0.05), Rajshahi (p<0.05), Rangpur (p<0.05) divisions. Poor mothers (p < 0.05) were more likely to go home delivery than rich mother. The mothers who were underweight (p<0.05), normal weight (p<0.05) and overweight (p<0.05) were more likely to perform home delivery than obese mothers. Mothers who were higher educated (p<0.05) were more likely to go delivery at private hospital/clinic than their counterparts. Mothers were more likely to have caesarean delivery at private hospital/clinic (p<0.05) than vaginal delivery. Mothers who spent more than 12000 taka for their last delivery were more likely to get delivery at private hospital/clinic than their counterparts. Currently working mothers had a greater chance to get delivery at hospital/clinic than non- working mothers (p<0.05) and mothers who were housewife were more likely to go private hospital/clinic for delivery (p<0.05) than mothers who were doing others work. It is necessary to take an action for making awareness about the bad effect of home delivery without skilled birth attendants among women in Bangladesh. Keywords: Place of delivery, Chi-square test, Multinomial logistic regression, Bangladesh Submission for Student Paper Competition? No

EC-040 Multinomial Logistic Regression Analysis of Factors Influencing Malnutrition of Non-Pregnant Married Women in Bangladesh: Evidence from Bangladesh Demographic and Health Survey-2014 Md. Minhazul Abedin1, Md. Enamul Haque2, Md. Sabiruzzaman2 Abu Sayed Md. Al Mamun2, and Md Golam Hossain2 1Department of Public Health, North South University, Dhaka, Bangladesh 2Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh

Under-nutrition and over-nutrition both are considered as malnutrition. Malnutrition is an important risk factor for various diseases. Body mass index (BMI) is used to measure malnutrition of a particular population. The purpose of this study was to determine the prevalence and factors associated of malnutrition among Bangladeshi non-pregnant married women in reproductive age. Data was extracted from the dataset collected by Bangladesh Demographic and Health Survey (BDHS) in 2014. The Chi-square test and multinomial logistic regression were used in this study. The present study revealed that the prevalence of

121

under nutrition and over nutrition among Bangladeshi married non-pregnant women in reproductive age was 18.6% and 24.4% respectively, it was dual burden for Bangladeshi women. Multinomial logistic regression model demonstrated that women who were living in Sylhet division (p<0.01), uneducated (p<0.01), poor (p<0.01), younger (age group in age≤20 years) (p<0.01), got marriage under 18 years old (p<0.01), early child bearing (p<0.01), living without husband (p<0.01), delivered at home (p<0.01), non-caesarean delivered (p<0.01) and currently breastfeeding (p<0.01) were more likely to get under nutrition than their counterparts. On the other hand, over nutrition was especially pronounced among women who were not living Sylhet division (p<0.01), lived in urban environment (p<0.01), came from rich family (p<0.01), older (age≥40years) (p<0.01), age at first marriage 18 years or above (p<0.01), currently living with husband (p<0.01), delivered at clinic/hospital (p<0.01), delivered by caesarean (p<0.01) and currently not breastfeeding (p<0.01). In this study, some modifiable factors were identified as predictors of malnutrition among Bangladeshi non-pregnant married women. Consequently, under nutrition can be reduced by promoting education level and improving the economic condition, but it is essential for making awareness about the bad effect of over-nutrition among Bangladeshi married women. Keyword: Nutritional status, logistic regression, nationally representative sample, Bangladesh Submission for Student Paper Competition? Yes

EC-041 Study on antenatal care among mothers in Bangladesh: a country based cross-sectional study Mosa. Shirina Akter1, Dolly Khatun1, Md. Belal Hossain1 Abu Sayed Md. Al Mamun1 and Md. Golam Hossain1 1 Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh

Antenatal care (ANC) is essential for pregnant women in order to ensure the best health conditions for both mother and child during pregnancy. The objective of this study was to determine the prevalence and associated factors related to ANC among Bangladeshi mothers. The data was collected by Bangladesh Demographic and Health Survey 2014. A total number of 4454 Bangladeshi mothers were considered as sample in this study. The multinomial logistic regression models were used in this study. This study revealed that only 31.97% mothers visited health care facility for 4 or more times to receive ANC during their pregnancy period. Multinomial logistic regression model demonstrated that uneducated, primary and secondary educated mothers were 11.26 times (p<0.001), 6.54 times (p<0.001) and 2.72 times (p<0.001) less likely to receive ANC service 4 or more times than higher educated mothers. Similar result was found when compared 1-3 ANC visits with 4 and more times. Respondents’ husbands’ education level and place of residence was also found to be a significant factor for influencing ANC visits of pregnant women. Underweight (p<0.001) and normal weight (p<0.001) mothers were more likely to receive no ANC visits during their pregnancy period than those who were obese mothers. Similarly, underweight and normal weight mothers were more likely to seeking 1-3 ANC visits 3.29 times and 2.41 times higher than obese mothers. A remarkable number of mothers do not receive ANC services in Bangladesh. Based on the study, some socio-economic and demographic factors have a strong and significant effect with receiving ANC services in Bangladesh. Our findings can help to Government and Non-government organizations in Bangladesh to improve police for increasing number of mothers receiving antenatal care services. Keyword: Antenatal care, logistic regression, nationally representative sample, Bangladesh Submission for Student Paper Competition? Yes

EC-042 Study on postnatal cares among mothers in Bangladesh: a country based cross-sectional study Mahmuda Khatun Eva1, Mosa. Shirina Akter1, Habibullah1, Nusrat Jahan Ema1 Abu Sayed Md. Al Mamun1, and Md. Golam Hossain1 1 Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh

Access to postnatal care (PNC) services has a great deal of impacts on major causes of infant death and significantly affects trends of mortality in a population. PNC is essential to save the life of the mother and newborn. The objective of this study was to determine the prevalence and associated factors of PNC among Bangladeshi mothers. In addition, identify the socio-economic, demographic and programmatic factors affecting not using PNC services. The data was collected by Bangladesh Demographic and Health Survey

122

2014. A total number of 4454 Bangladeshi mothers were considered as sample in this study. The multinomial logistic regression models were used in this study. This study revealed that 64.84% mothers received PNC services within two months after delivered their babies. χ2-test demonstrated that division, educational level, mode of delivery, husbands’ education level, respondents’ currently working status, number of s visits, wealth index, total ever born children, age at first birth, respondents’ body mass index, husbands' occupation, respondents' occupation, age at first marriage, type of place of residence were significantly associated with PNC among Bangladeshi mothers. Multiple binary logistic regression analysis was utilized to find the effect of socio-economic, demographic and anthropometric factors on number of receiving PNC services among Bangladeshi mothers. After controlling the effect of other factors, this model demonstrated that Bangladeshi mothers were living in Sylhet were less likely to receive PNC than Chittagong (p<0.05) and Rangpur ( p<0.05) divisions. Education had positive effect on postnatal care showing that Bangladeshi mothers whose husband having higher education (p<0.05) were more likely to receive postnatal care than who were no formal education. It was found that caesarean delivered mothers ( p<0.05) more likely to receive PNC services than vaginal delivered mothers. Bangladeshi mothers who had received 4 or more times ANC services ( p<0.05) they were more likely to receive postnatal care than their counterparts. It was also found that urban mothers were more likely to get PNC services than mothers who lived in rural environment. The main objective of this study is to understand why some mothers in Bangladesh do not receive PNC services. It is found that a remarkable number of mothers do not receive PNC services in Bangladesh. Mothers and their husbands’ education appear to have a strong and significant association with receiving PNC services in Bangladesh. Community based intervention and regular home visits by health care providers could enhance care for women and newborns including delivery of specific health messages. Counseling could be integrated during antenatal care visits to increase the postnatal care service further. Our findings can help to Government and Non-government organizations in Bangladesh to improve police for increasing number of mothers receiving antenatal and postnatal care services, consequently to reduce maternal and neonatal mortality. Keyword: Postnatal care, logistic regression, nationally representative sample, Bangladesh Submission for Student Paper Competition? No

EC-043 Factors influencing early menopause in Bangladesh Md. Abdul Awal1, Farzina Parvin2, Md. Abdul Wadood2, Jarin Sazzad3 Abu Sayed Md.Al Mamun2, Md. Golam Hossain2 1Department of Public Health, Varendra University, Rajshai, Bangladesh 2Health Research Group, Department of Statistics, University of Rajshahi, Bangladesh, 3Department of Microbiology, Rajshahi Medical College, Rajshahi-6000

The prevalence of non-communicable diseases is a challenging problem among menopausal women in Bangladesh, where suffering from chronic diseases after menopausal age is very common. Early menopause is a risk factor for health among Bangladeshi married women. In Bangladesh, the early menopausal rate has been increasing with increasing the household wealth quintile and women education. The aims of this study were to determine the prevalence and associated factors of early menopause and find its most important predictors. In this study, the secondary data was used that was extracted from the 2014 Bangladesh Demographic and Health Survey (BDHS-2014). BDHS-2014 collected data from overall Bangladesh using two stage stratified cluster sampling method. Descriptive statistics, Chi-square test, binary logistic regression and stepwise logistic models were used in this study to determine prevalence, associated factors and effect of selected factors on early menopause respectively among Bangladeshi married women aged 30- 49 years. This study revealed that the prevalence of early menopause among Bangladeshi married women was 17.9%, and women living alone [AOR=1.397, 95% CI: 1.084, 1.801; p<0.01)], Muslims [AOR=1.503, (95% CI: 1.108, 2.037) p<0.01] and early marriage (age<18 years) [AOR=0.788, (95% CI: 0.605, 1.026); p<0.05] were the most influential factors. The prevalence of early menopause is high among women aged 30-49 years in Bangladesh, and some modifiable factors were found to be the important predictors. Our findings can be considered for achieving target regarding mothers’ morbidity, an important component of Sustainable Development Goals by 2030. Key words: Early menopause, BDHS–2014, Non-pregnant women, Logistic regression. Submission for Student Paper Competition? No

123

EC-044 Pre-eclampsia and associated factors among pregnant women in Rajshahi Medical College Hospital, Rajshahi, Bangladesh Jannatul Ferdous3,Md. Abdul Awal1, Jarin Sazzad,2 Md. Abdul Wadood3 Abu Sayed Md. Al Mamun3, Md. Golam Hossain3 1Department of Public Health, Varendra University, Rajshai, Bangladesh, 2Department of Microbiology, Rajshahi Medical College, Rajshahi-6000 3Health Research Group, Department of Statistics, University of Rajshahi, Bangladesh.

This cross sectional type of descriptive study was carried out among the patients of pre-eclampsia of pregnancy admitted in Rajshahi Medical College Hospital with a view to observe pre-eclampsia in pregnancy and associated risk factors. The sample size was 200, which was selected purposively. The mean age of the respondents was 26.16 ± 4.13 years. The average monthly family income was BDT 13540.00 ± 5235.02. About 65% of the respondents had irregular Ante Natal Care (ANC), 95% of them had systolic hypertension and 96% had diastolic hypertension. It was also observed that pre-eclampsia of pregnancy was significantly associated with age of the respondents (p<0.001). This study provided some important information which might help the concerned people to take appropriate measures and might be the basis for further in-depth study on this issue. Keywords: Chi-square test, pregnant women, Rajshahi, Bangladesh. Submission for Student Poster Competition? Yes

EC-045 The Impact of Lending Interest Rate on Nonperforming Loans Rate in the Banking Sector in Bangladesh Dr. Shamim Ara1 and Mohammad Mohidul Islam2 1Deputy General Manager, Statistics Department, Bangladesh Bank, Head Office, Dhaka-1000, E-mail: [email protected] 2Joint Director, Statistics Department, Bangladesh Bank, Head Office, Dhaka-1000, E-mail: [email protected]

This study is a modest attempt to investigate the impact of lending interest rates on the NPL rate in the banking sector in Bangladesh. The empirical analysis employed quarterly time-series data of NPL rate and lending interest rate of the overall banking system of Bangladesh from 2000:Q1 to 2019:Q1. This paper examined the short-run and long-run relationships between NPL rate and lending interest rate using modern time series co-integration technique based on autoregressive distributed lag (ARDL) bounds test method. The findings of the study revealed that the long-run relationship exists between the variables and the short- run dynamic adjustment has taken place in a consistent manner only in the NPL equation. The conclusion of the research also suggested that the monetary regulator should take the right measures for handling the country's NPLs situation properly. Keywords: NPLs, Lending Interest Rate, Co-integration, ARDL and Bangladesh.

EC-046 Prevalence of depression and its correlates among University students in Bangladesh Al-Hasib Billah1 and Mohammad Ali1 1Statistics Discipline, Khulna University, Khulna-9208, Bangladesh

Background: Depressive symptoms is now a worldwide concern of mental health issue, ranking third after cardiac and respiratory diseases as a major cause of infirmity with an indication to recommend that university students are at higher risk of depression, despite being a socially advantaged population, but the reported rates have shown wide variability across locations. Objective: The intentions were to determine the prevalence of depression among University students in Bangladesh and also find out some risk factors linked to depression. Methods: A cross sectional study was carried out with sample size 600. These 600 participants were selected randomly from three prominent universities in Bangladesh by stratified random sampling. The questionnaire included the socio-demographic characteristics, Sleep quality index (PSQI) & Patient health questionnaire

124

index (PHQ-9). The data were analyzed by using percent (%) ratios, chi-square and Binary logistic regression analysis with a significant value of P<0.05. Results: Out of 600 students, 394 (65.7%) were males and 206 (34.3%) were females. Results found that 76.2% (n=457) students are suffering from depression. The prevalence of depression was found more obvious in case of male students 73.4% (n=289/394) than female 81.6% (n=168/206). In addition gender, Father's Education, social acceptance increasing thoughts, ever insulted by close one and sleep quality variables have been found as significant factors. Conclusions: It is believed that university graduates are the future leader and progressors of the nation. The higher percentage of depression is an alarming issue for the students studying in a university environment as well as for general people. The authority should take some curative measurements to lessen the occurrence of depression such as counseling & awareness campaigning.

EC-047 Estimating Market Value-at-Risk of DSEX Index Based on GARCH Models. Md. Monimul Huq1, Md. Ayub Ali1 1Department of Statistics, University of Rajshahi, Bangladesh, E-mail: [email protected]

This study conducted the empirical investigation for the volatility of Dhaka stock Exchange Broad index (DSEX) in Bangladesh. In light of the up-to-date universal monetary crisis and the continuing independent debt crisis, accurate measuring of market losses has become a very current issue. One of the leading standard risk measures is Value-at-Risk (VaR). A set of symmetric and asymmetric GARCH family models based on several error distributions were applied on Dhaka Stock Exchange DSEX Index from January 27, 2013 to June 30, 2019 for estimating and forecasting the market Value-at-Risk of the index. The most adequate GARCH family models for estimating volatility in the Dhaka stock exchange was found to be as the GJR- GARCH (1,1) model with Student’s-t error distribution. The model selection accepts that the GJR-GARCH (1,1) with Student’s-t error distribution model because it improves the log likelihood and the AIC, BIC and HQ criterion. GJR-GARCH (1,1) model with Student’s-t error distribution was allowed by Kupiec test with 99% of confidence level. The proposed VaR model would help the investors in their evolving investment marketplaces. Keywords: Value-at-Risk, Volatility, GJR-GARCH model, DSEX.

EC-048 Automated Detection and Classification of Diabetes Disease Based on Machine Learning Approach in Bangladesh Md. Merajul Islam1, Md. Jahanur Rahman1, Dulal Chandra Roy1, Md. Maniruzzaman1, 2 1Department of Statistics, University of Rajshahi, Rajshahi, Bangladesh 2Statistics Discipline, Khulna University, Khulna, Bangladesh

Background and Objectives: Diabetes mellitus has been recognized as a continuing health challenge for the twenty-first century, both in developed and developing countries including Bangladesh. Diabetes is a chronic disorder that can affect the body's main organs and leads to heart disease, stroke, amputation, kidney failure, blindness and early death. Early detection of risk factors is important to prevent the associated diseases of diabetes. When recently developed machine learning based automated models have attracted increasing attentions for detection and classification of diabetes, no significant research works are found in Bangladesh to detect and classify diabetic disease using these methods. For the first time, an attempt has been made in this study to use machine-learning methods for automated detection and classification of diabetes based on Bangladesh Demographic Health Survey (BDHS 2011) data. Materials and Methods: The diabetes dataset, taken from Bangladesh Demographic Health Survey (BDHS 2011) data. The dataset contained sixteen variables and 1,569 respondents with 127 diabetic. Risk factors are identified using two feature selection methods namely independent t-test for continuous variables and chi- square test for categorical variables. Six ML based classifiers as support vector machine (SVM), random forest (RF), linear discriminant analysis (LDA), logistic regression (LR), k-nearest neighborhood (k-NN), bagged classification and regression tree (Bagged CART) have been compared to identify the best model based on the highest classification accuracy. Performance analysis is done using Accuracy (ACC), Sensitivity (Sen), Specificity (Spec), Mathew Correlation Coefficient (MCC) and Area Under the Curve (AUC).

125

Results: This study showed that residence, electricity, wealth index, age, education, working status, arm circumference, take medicine, weight and BMI were found to be significant risk factors for diabetes. Among six classifiers, Bagged CART provided the highest ACC and AUC of 94.3% and 0.941 compared to others for detecting and classifying the diabetes disease. Conclusion: Our study showed that Bagged CART method anticipated a very supportive computational resource for detection and classification of diabetes disease using BDHS (2011). It will be very helpful the physicians and doctors to make a decision to control diabetes disease in Bangladesh.

EC-049 Relationship among Mental Health Variables: Comparison between Facebook Users and Non-users in Khulna University Kamrul1, Mohd. Mohaiminul1, Sawda1 and Md. Menhazul1 1Statistics Discipline, Khulna University, Khulna

Facebook was founded in 2004 and has become one of the most popular social networking sites. Nowadays, over 1.5 billion people are connected with the social networking site (SNS) Facebook. Although Facebook is used to connect and stay in touch with friends and family and also to maintain relationships, there may be situations where an individual excessively uses this site. It's likely to take a toll on an individual’s emotional well-being if one depends on it for social connections. The aim of the present study was to investigate the relationship and differences in various mental health variables between Facebook users and non-users in Khulna University. The data of 365 participants (306 Facebook users and 59 Facebook non-users using simple random sampling) were collected. In our study, we computed Box’s M test for homogeneity, Hoteling’s Trace to see the presence of the significant effect and MANOVA (multivariate analysis of variance) to see the significant mean difference. The collected data was analyzed using SPSS software. In our study, we find that Facebook users score significantly lower on life satisfaction (SWLS) and subjective happiness (SHS) than Facebook non-users. Facebook non-users have significantly lower values of depression symptoms than Facebook users. In both groups, DASS-21: Depression, DASS-21: Anxiety, DASS-21: Stress symptoms, on the one hand, subjective happiness (SHS) and life satisfaction (SWLS), on the other hand, are negatively correlated. Here, we found significant differences between Facebook users and Facebook non-users regarding some mental health variables. The present results indicate that Facebook non- users have higher values of positive variables protecting mental health compared to Facebook users. These findings are of particular no interest in using Facebook as the use of Facebook has a negative impact on the mental health of many students of Khulna University. Keywords: Facebook, Mental Health, Multivariate Analysis, MANOVA.

EC-050 Performance Evaluation of Classification Learning Models for Wisconsin Breast Cancer Data Repository Sutapa Dey Barna1, Md. Salauddin Khan1 1Statistics Discipline, Khulna University, Khulna, Bangladesh, [email protected] 1Statistics Discipline, Khulna University, Khulna, Bangladesh, [email protected]

Breast cancer is one of the most frequently occurred diseases for women in the modern era. So, primary detection of breast cancer will give a chance to early medication of the women. As there are difficulties in accurate analysis of some mammogram images, some concretely classifiers are being used for biomedical analysis. In this study, six different classification models such as Logistic regression, Linear discriminant analysis (LDA), Quadratic discriminant analysis (QDA), K-nearest neighbor (KNN), Naive Bayes (NB) and Artificial neural network (ANN) are used to predict the breast cancer risk behavior. To classify breast cancer data, training and test data set are carried out four proportions such as 60:40, 70:30, 75:25 and 80:20. Finally, the overall performances of the six models are evaluated using sensitivity, specificity, accuracy, and precision. Data that contains 683 patients, are collected from the UCI machine learning repository. The result shows that the performance of ANN in 75:25 proportion as accuracy 97%, sensitivity 94%, specificity 96% and precision 98% which is best compared to other proportions, among the selected models. Moreover, ANN performs equally well, especially in 60:40 ratio, though NB having around 95% accuracy. Overall, ANN classifies breast cancer promptly compared to other classification models which are used in our study. Keywords: Breast cancer, data mining, classification, knowledge retrieval and ANN.

126

EC-051 Parents’ mental illness and child behaviour: a mediation analysis Md Jamal Uddin1,2, Merete Nordentoft3,4, Claus Thorn Ekstrøm2 1. Department of Statistics, Shahjalal University of Science and Technology, Sylhet, Bangladesh 2. Section of Biostatistics, Department of Public Health, University of Copenhagen, Denmark 3. Copenhagen University Hospital, Mental Health Centre Copenhagen, Denmark 4. The Lundbeck Foundation Initiative for Integrative Psychiatric Research (iPSYCH), Copenhagen and Aarhus, Denmark

Children of parents with severe mental disorders, e.g., schizophrenia (SZ), bipolar disorder (BP), are exposed to numerous environmental risk factors and facing several problems like cognitive, social development, and academic functioning. These problems may have related to their home environments where they grownup and/or their parents’ education. We aimed to investigate the association of parents’ mental disorders and child behaviour when home environments and/or parents’ education act as mediators. We used data from the nationwide Danish High Risk and Resilience Study – VIA7. The study sample was 522 in which 202 7-year-old children living with their parents diagnosed with SZ, 120 children living with their BP parents and 200 control children living with their parents who did not have a diagnosis of SZ or BP. The outcome variable was considered as child behaviour checklist school-age version (CBCL). We analysed data using different statistical models including parallel and serial mediation models. Our mediation analyses showed that the total effects of BP and SZ as compared to the control group on the CBCL was 6.50 [2.15,10.85] and 9.96 [6.22,13.69], respectively. When considering the mediator home environment only, it could account for 43% of the total effect of the parents’ mental disorders on the CBCL. When considering both mediators, there was a significant indirect effect of parents’ mental disorder on the CBCL through home environment and parents’ education. We conclude that an adequate home environment and/or parents’ education are important mediating factors for child development, particularly for child behaviour. Keywords: Parents mental disorders, child behaviour, home environments, parents’ education, mediation analysis Submission for Student Paper/Poster Competition? No

EC-052 Impact of Depression, Anxiety and Stress on daily average Sleeping duration: A case study among Varendra University Students Md. Nadim Sarwar1, Ahammad Hossin2& Prof. Dr. Md. Ayub Ali3 1B.Sc. (Hon’s) student, Dept. of CSE, Varendra University, Email: [email protected] 2Assistant Professor, Dept. of CSE, Varendra University, Email: [email protected] 3Professor, Dept. of Statistics, University of Rajshahi, Email: [email protected]

Now a day’s depression, anxiety and stress (DAS)of university students is a big concern over the world. The main objective of this study was to detect the impact of depression, anxiety and stress with corresponding to the average daily sleeping duration among the students of Varendra University. This Study deals with the primary data of which was collected from face to face questionnaire survey through simple random sampling without replacement method from 5134 students under the 11 departments of university. Evaluation of depression, anxiety and stress were calculated using DASS-42 structured questionnaires. The statistical analysis like frequency distribution, cross tabulation, relevant statistical tests and correspondence analysis were conducted using IBM SPSS of version 23 and in some of the cases, python programing language were used. To detect the significant association between DAS and daily average sleeping duration, Chi-square test, Linear-by-Linear Association test, Somers’d, Kendall's tau-b and Gamma symmetric measure were also analyzed. This study reveals that majority of the students of VU follows normal depression was 39.1%, moderate anxiety was 35.2% and normal stress was 53.9%.The statistical tests suggest that there are statistically significant association between depression and daily average sleeping duration, between stress and daily average sleeping duration, but not between anxiety and daily average sleeping duration. As per the normal level, the probability of normal depression and normal stress are 0.879 and 0.881 respectively among the respondents who slept 5 hours to 9 hours. From these results we can say that,5 hours to 9 hours average sleeping duration can be able to reduce the severity of depression and stress. To avoid mild, moderate, severe and extremely severe depression and stress, we should take a sound sleep at least 5 to 9 hours on daily average basis. Keywords: Depression, Anxiety, Stress and Average Sleeping Duration

127

EC-053 Awareness of Anemia among Rural Pregnant Women in Bagerhat District of Bangladesh Md. Abu Borhan1 and Md. Ayub Ali2 1 MPSc in Applied Statistics, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh 2 Professor, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh

Background: Anemia in pregnancy is a decrease in the total red blood cells (RBCs) or hemoglobin in the blood during pregnancy or in the period following pregnancy. The condition of having a lower-than-normal number of red blood cells or quantity of hemoglobin. Anemia diminishes the capacity of the blood to carry oxygen. Patients with anemia may feel tired, fatigue easily, appear pale, develop palpitations, and become shortness of breath. Objectives: The purpose of the present study was to find out the awareness of anemia among rural pregnant women in Bagerhat district of Bangladesh Materials and Method: A sample of 29 pregnant women (PW) from a total of listed 111 women from three upazilas of Bagerhat district was considered for assessing the awareness of Anemia. Those three upazilas were taken at random first from the nine upazilas of Bagerhat district. Data on different variables were collected directly from the selected women through a prescribed questionnaire. Descriptive statistics e.g., maximum, minimum, mean, standard deviation, skewnwss, kurtosis, etc of the variables together with their standard error of estimates were considered for analyzing sample characteristics of the study. The relationship between two nominal variables is assessed by cross tabulation with test statistics Phi and Cramer's V. Results: About 86% pregnant women have no idea about anemia and also their causes. All respondents fell weakness that indicates they have the symptom of anemia. The phi and cramer's V imply that the relationship between the knowledge about anemia and sources of information is highly significant ( p= 0.000). Among the awarded (14%) women, about 92% have got the information from service provider. Among the population the contribution the service provider is .379. Probability of unknown is 0.586 indicating much population in Bagerhat district does not know what is anemia. Conclusion: Probability of getting information of anemia from mother is zero indicating very recently service providers have started their program in Bagerhat district. Therefore, this program should be continued until the probability of getting information will be closed to 1. Recommendation: Government as well as the NGOs should continue & enhance the present awareness program in Bagerhat district. Keywords: Awareness, Anemia, Pregnant women, Correspondence Analysis, Phi Correlation and Cramer's V

EC-054 Association of IL-6 Gene rs1800796 Polymorphism with Cancer Risk: A Meta-Analysis Md. Harun-Or-Roshid1, Md. Borqat Ali1, Jesmin2* and Md. Nurul Haque Mollah1*, 1Bioinformatics Lab. (dry), Department of Statistics, University of Rajshahi, Rajshahi-6205; 2Department of Genetic Engineering and Biotechnology, University of Dhaka, Dhaka-1000, Bangladesh; *Correspondence: Mollah ([email protected]); Jesmin ([email protected])

Interleukin-6 (IL-6) gene polymorphisms are a crucial functional marker in human body. Various genetic studies have shown that IL-6 genes are significantly associated with major diseases and various cancers. In this study, we want to confirm the association between the IL-6 gene polymorphism (rs1800796) and the cancer risk through a meta-analysis. For the meta-analysis of Genome Wide Association Study (GWAS) of IL-6 gene (-572G/C) polymorphism , 27 case-control study with sample size 24138, through searched the relevant studies published up to February, 2019 and performed an efficient searching strategy from PubMed, PubMed Central, Web of Science, Google Scholar and the relevant Biological literature based online databases. To investigate the association we compared the allelic and genotypic case-control frequency via Odds Ratio (OR), 95% confidence interval (CI) and some other statistical test. In the study of rs1800796 SNP, we found a significant association for increasing risk (CG vs. CC+GG: OR = 1.12, 95% CI= 1.01 – 1.23, p = 0.0288) for overall cancer, and subgroup analysis suggested that lung and stomach cancer increasing risk as well as a protective role for prostate cancer in Asians. These findings suggest that IL-6 gene polymorphisms may appraise as a genetic biomarker for cancer risk. Keywords: IL-6 gene, Cancer, GWAS, Case-Control studies, Meta-analysis

128

EC-055 A Statistical Study on Initial Breastfeeding among Mothers in Rajshahi District, Bangladesh Ummay Ayesha1, Md. Nurul Islam1, Md. Ripter Hossain1, Abu Sayed Md. Al Mamun1 and Md. Golam Hossain1 1Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh

Breastfeeding is considered as one of the most important factors for growth and development of infants. Early initiation of breastfeeding (EIBF) is referred as providing the breast milk to the newborn infant within one hour of birth which ensures that the infant receives the colostrums. The aim of this study was to identify the risk factors associated with initial breastfeeding among mothers in Rajshahi district, Bangladesh. The data was collected from mothers in Rajshahi district who had at least one child aged 6-24 months from January to March, 2019. Multistage random sampling method was used for selecting our sample. Frequency distribution, Chi-square test and binary logistic model was used in this study to find the associated factors of initial breastfeeding among mothers. The prevalence of initial breastfeeding among mothers in Rajshahi district, Bangladesh was 88.4%.Out of 450 mothers, near half of the mothers (47.5%) delivered their child in public hospital where 10.7 % and 41.8% delivered in home and private hospital respectively. The higher number of mother’s (68.1%) monthly income was below15000 Taka where 8.1% above 25000 Taka and 3.8% income level are between 15000 Taka and 25000 Taka. Chi-square test provided that mode of delivery, place of delivery, respondents’ age, respondents’ and their husbands’ education level, respondents’ nutritional status, family monthly income, respondents’ age at first birth were significantly associated factors of initial breastfeeding among mothers. Binary logistic model demonstrated that mothers who delivered at home was more likely to provide their initial breast milk to their infant [AOR=0.090, 95% CI:0.010-0.794; p<0.05] than mothers delivered at private hospitals. It was noted that mothers living in poor income who were more interested to provide their initial breast milk to their infants than middle incoming family [AOR=0.097, 95% CI: 0.030 -0.315; p<0.01] and rich family [AOR=0.130, 95% CI: 0.039-0.437; p<0.01]. We found that primary or uneducated husbands’ wives had more chance to give their initial breast milk to their infants [AOR=0.167, 95% CI: 0.035 -0.792; p<0.05] than higher educated husbands’ wives. The nutritional status of mothers was an important predictor of initial breastfeeding, and it was observed that under nourished mothers were more likely to provide their breast milk to their infants than healthy [AOR=0.108, 95% CI: 0.013-0.867; p<0.05] and over nourished mothers [AOR=0.072, 95% CI: 0.008- 0.692; p<0.05]. This study identified several socio economic factors that were associated with initial breastfeeding, and hope that this information will help the government and other policy makers to take proper initiative for promoting the awareness of initial breastfeeding in Bangladesh. Key Words: Initial breastfeeding, Rajshahi, Bangladesh, logistic regression Submission for Student Paper Competition? Yes

EC-056 Statistical Prediction of Lysine PTM Sites Mapping on Homo Sapiens under the CKSAAP Encoding Samme Amena Tasmia*, Md. Merajul Islam, Md. Harun-Or-Rashid, Fee Faysal Ahmed, Md. Parvez Mosharaf, Md. Hadiul Kabir and Md. Nurul Haque Mollah* Bioinformatics Lab., Department of Statistics, Rajshahi University, Rajshahi-6205, Bangladesh; *Correspondence: Mollah([email protected]); Tasmia([email protected])

Post-translational modification (PTM) refers to the covalent and enzymatic modification of proteins which plays a key role in protein conformation regulation and cellular function control. Identification of Lysine PTM sites can facilitate our understanding about the molecular mechanism accurately. However, traditional experimental approaches of PTM site predictions are labor-intensive and time-consuming. In this context, a more accurate computational method for predicting Succinylation sites is an urgent issue which can be useful for drug development. In this work, we developed a novel Succinylation site predictor called Succinpred, which is constructed including CD-HIT (removing 40% identity), 5-fold cross-validation, CKSAAP encoding, 1:2 ratio of the positive vs. negative samples, and AdaBoost. The performance of this method was measured with an accuracy of 90.2%, a MCC (Matthew Correlation Coefficient) of 81.9% and AUC (Area under the ROC Curve) of 0.978 using 5-fold cross validation on training dataset and an accuracy

129

of 89.5%, a MCC of 80.4% and AUC of 0.948 on independent dataset. The proposed predictor also performs much better than the other existing predictors. Keywords: Succinylation site, Feature Extraction, Sequence Analysis, CKSAAP and Machine Learning.

EC-057 Somatotype, BMI, Motor Fitness as Performance Indicators of the Adolescent Athletes: A Review A.H.M. Rakibul Mawla1 and Md. Ayub Ali2 1 PhD Fellow, Department of Statistics, University of Rajshahi and Associate Professor, Department of Statistics, University of Chittagong Email: [email protected] 2 Professor, Department of Statistics, University of Rajshahi E-mail: [email protected]

Studies on Somatotype, BMI and Motor Fitness are very important for talent hunt in sports. Policy maker and sports administrators as well as the managers and coaches must have the idea about the potentiality of the athletes involved in various events of sports. The purpose of the present study was to review the literature relevant to the topic keeping in mind the necessity and importance of the knowledge regarding the potentiality of the athletes. These types of works are mostly available in Kalyani University, Bengal, India. However, the statistical methods that they have been applying need to be improved accordingly. On the other hand, Bangladesh is now improving their sports ability, but their performance potentiality needs to be scientifically assessed. The present review study will help us to enhance this assessment. Keywords: Somatotype BMI Motor fitness Performance indicators Adolescent athletes

EC-058 Identifying factors associated with course drops among students in SUST Mohammad Shahidul Islam, Ruhul Amin, Md. Wahiduzzaman khan, Md. Ahadur Rahman, Md. Najmul Islam and S. M. Khurshid Alam* Department of Statistics, Shahjalal University of Science & Technology, Sylhet, Bangladesh *Corresponding author: S.M. Khurshid Alam, Email: [email protected]

Course drop means the situation where a student fails in a course. It can happen because of poor performance of the student in exam paper. However, sometimes students do it purposefully, out of fear that they will get a bad grade, in a hope to do better, next time he/she participates in the exam. The aim of this study is to fit a model of number of course drops and identify the factors associated with it among the students in Shahjalal University of Science and Technology (SUST). Primary data from 471 students, currently studying in different departments of SUST were collected. Negative Binomial Regression model was employed for proper modeling. It was found that students’ higher secondary institutes and background, faculty, residence, gender and any family member having chronic/acute disease are some of the factors highly influence the occurance of course drop. The finding can be utilized, and can be extended for other similar institutions, to undertake proper steps and measures to reduce the course drop among the students. Keywords: course drop, SUST, negative binomial regression

EC-059 Status of Sustainable Development Goal 1 within the Student of Varendra University Md. Nadim Sarwar1, Ahammad Hossin2 and Md. Ayub Ali3 1B.Sc. (Hon’s) student, Dept. of CSE, Varendra University, Email: [email protected] 2Assistant Professor, Dept. of CSE, Varendra University, Email: [email protected] 3Professor, Dept. of Statistics, University of Rajshahi, Email: [email protected]

Now-a-days, poverty is the main problem over the world. The main topic of this study was to investigate the status of sustainable development goal 1among of students of Varendra University, Rajshahi. A sample of 358 students under 11 departments was considered for this study. Descriptive statistics were conducted using IBM SPSS of version 23 here. This study reveals that there are 30.2% student’s family living under poverty line, followed by 28.2% living on low income, 37.4% living on lower middle income, 4.2% living on higher middle income and no family living on higher income. Most of the student from pure family have got admission in private university like Varendra University, Rajshahi through scholarship or waiver for their brilliant previous results. Government should take necessary actions to remove the poverty from everywhere

130

of our country and until that to increase scholarship/subsidies to achieve the sustainable development goal 4(Education for all). Keywords: Goal 1, Goal 4, Poverty

EC-060 Adolescent Obesity and Mental Health Status in Asian Countries: A Systematic Review Papiya Roy1, Premananda Bharati2, Suman Chakravarty3* 1Assistant Professor, Dept. of Anthropology, Government General Degree College, Keshiary, Paschim Medinipur, West Bengal, India, Pin- 721135, Email ID: [email protected] 2Retired Professor, Biological Anthropology Unit, Indian Statistical Institute, 203, B.T. Road, Kolkata- 700108 , India, Email ID: [email protected] 3Assistant Professor, Dept. of Anthropology, Mrinalini Datta Mahavidyapith, Birati, Kolkata – 700051, India, Email ID: [email protected]

Adolescence is a critical period for developing weight related problems and also related to depression. According to National Mental Health Survey (NMHS) 2015-16, it has been declared that there is no health without mental health. This study is an attempt to do the systematic review of literature on adolescent obesity related to mental health status in the Asian countries. Studies were identified from the electronic databases like Pubmed, Science Direct, Psych INFO, Google Scholar etc. A total of 252 articles, 160 articles were extracted out of which only 8 articles fulfill the eligibility criteria. Age group varies from 10 to 19 years. For physical health status, Body Mass Index (BMI) is the primary indicator and for mental health status BSQ, BSD, BDI and RSES being the primary indicator. Result shows that a significant positive relationship has been found between BMI and body shape measure. Studies also suggest that the common mental health problems among the obese adolescents are anxiety disorders, depressive disorders and behavioural disorders. The key factors are responsible for these are age, sex, parent’s education, parent’s occupation, family income, academic stress, body shape dissatisfaction, body image perception and self esteem. Key words: Obesity, Mental Health status, Adolescents, Asia

EC-061 Bioinformatics analysis of differentially expressed genes in Mouse Striatum using RNA-Seq data Bandhan Sarker1, Suman Khan1, Priyanka Bosu1, Md. Nurul Haque Mollah2,* and Md. Matiur Rahaman1,* 1Department of Statistics, Faculty of Science, Bangabandhu Sheikh Mujibur Rahman Science and Technology University, Gopalganj-8100, Bangladesh. Email: [email protected]; [email protected]; [email protected]; [email protected] 2Bioinformatics Laboratory, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh. Email: [email protected]

Two most commonly used inbred mouse strains in neuroscience research are C57BL/6J (B6) and DBA/2J (D2). However, the only currently available mouse genome is based entirely on the B6 strain sequence. This study aims to complement the differentially expressed genes (DEGs) of B6 and D2 samples, gene-gene interaction network, identify their GO functions, KEGG pathways and downstream analysis. The gene expression profile of GSE26024 from Gene Expression Omnibus database, including 10 B6 and 11 D2, was deeply analyzed. DESeq2, EdgeR and limma R tool were used to screen the differentially expressed genes (DEGs) between B6 and D2 samples. Search tool for the Retrieval of Interacting Genes and Cytoscape software were used to visualize the protein–protein interaction of these DEGs. Moreover, Gene Ontology analysis and Kyoto Encyclopedia of Genes, Genomes pathway analysis and downstream analysis were performed for screening DEGs. Total of 736, 757 and 530 DEGs for DESeq2, EdgeR and limma R, respectively, had identified 37, 48, and 31 up-regulated genes respectively. Using gene interaction network analysis revealed top 50 high-degree hub nodes of DEGs for among those three methods and identified 21 common hub genes, whereas three genes are up-regulated. Then those 21 common hub genes were analyzed by GO, KEGG and downstream analysis. Using downstream analysis, k-means and hierarchical clustering found Bdkrb2, Aplnr and Ccl28 genes belong to the same group. Taken together, DEGs, especially the hub genes Bdkrb2, Aplnr and Ccl28, might be the driven genes in inbred mouse strains. Keyword: RNA-Seq data, differentially expressed gene, GO, KEGG, inbred mouse strains. Submission for Student Paper? Yes.

131

EC-062 Trend in malnutrition among married women in Bangladesh: Urban and rural difference Md. Kamrul Islam, Md. Sabiruzzaman and Md Golam Hossain Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh [email protected], [email protected], [email protected]

Body mass index (BMI) for adults is used to measure nutritional status of a particular population. Nutritional status is an important indicator of health status of a nation. The change of BMI over time may provide useful information about changes in the level of public health and reflect the general living environment of a given population. This information is crucially important for developing countries like Bangladesh where health and medically related reforms are being actively implemented. The aim of this study was to observe the change in the category of BMI (underweight, overweight and obese) over time among Bangladeshi married non-pregnant women in reproductive age. Data was extracted from Bangladesh Demographic and Health Survey (BDHS) 1999-2000, 2004, 2007, 2011 and 2014. It was observed that the prevalence of underweight women have been decreasing during the last one and half decades (1999-2000 to 2014). However, the prevalence of overweight and obese among Bangladeshi women has been increasing during the same time period. It was noted that the prevalence of underweight among rural women was significantly higher than that of urban women for each survey period (p<0.01). The number of overweight and obese (percentage) women in urban area was significantly (p<0.01) higher than that in rural area. This study supports that over nourished (over weight and obese) is increasing with increasing household wealth index in Bangladesh. It is suggestive that the health authorities of Bangladesh should improve policy for making awareness among especially married women about the importance of balanced diet and healthy lifestyle. Keyword: Over nutrition, under nutrition, married women, urban rural difference, and proportion test. Submission for Student Paper Competition? No

EC-063 Transformed Linear Discriminant Analysis and its Application in Gene Expression Data Classification Priyanka Bosu, Bandhan Sarker, Suman Khan and Md. Matiur Rahaman* Department of Statistics, Faculty of Science, Bangabandhu Sheikh Mujibur Rahman Science and Technology University, Gopalganj-8100, Bangladesh. Email: [email protected]; [email protected]; [email protected]; [email protected].

Gene expression data analysis plays the significant role in identifying biomarkers (genes)/patients of the species because of the advanced generation of genomic technologies. Classification and clustering methods are essential to classify/predict the genes/patients that is highly correlated for prognosis diseases, imposing classes of the diagnosed diseases, as well as analyzing and explaining molecular level for detecting potential biomarkers that are linked to causing specific diseases in organism. Statistical approaches employed for the classification of gene-expression data are based on either a continuous scale (e.g. microarray data) or need to fulfill the normality assumptions. However, gene-expression data often contaminated by outliers (noise), which are affect data analysis results. Therefore, these statistical algorithms can be applied with proper modifications. In this paper, we proposed a logistic transformation based Linear Discriminant Analysis (LDA) method in the context of gene expression data analysis. A comprehensive simulated gene expression and a real data analysis were carried out to evaluate the performance of our proposed method. The proposed LDA provides much better results than the classical LDA when datasets are contaminated by the noise i.e. outliers. Otherwise, it keeps equal performance. Keywords: Gene expression, Microarray, LDA, Outliers, Transformation. Submission for Student Paper Competition? Yes

132

EC-064 Protein-Protein Interaction Prediction in Arabidopsis Thaliana Using Machine Learning Algorithms Fee Faysal Ahmed1, Mst. Shamima Khatun2 and Md. Nurul Haque Mollah3* 1Department of Mathematics, Jashore University of Science and Technology, Jashore-7408, Bangladesh. Email: [email protected] 2Department of Bioscience and Bioinformatics, Kyushu Institute of Technology, 680-4 Kawazu, Iizuka, Fukuoka 820- 8502, Japan. Email: [email protected] 3Bioinformatics Lab., Department of Statistics, Rajshahi University, Rajshahi-6205, Bangladesh. . Email: [email protected] *Correspondence: Mollah ([email protected])

Arabidopsis thaliana is a standard model organism in plant biology and genetics. To understand the molecular structures and functions of the Arabidopsis thaliana sequence, identification of protein-protein interactions (PPIs) is very important. PPIs indicate the physical contacts between two or more protein molecules that occur due to the biochemical reaction steered by electrostatic forces including the hydrophobic effect. Protein rarely works without PPIs. Many molecular processes within a cell are carried out by their PPIs. The irregular and unusual PPIs are the basis of multiple aggregation-related diseases, such as Creutzfeldt–Jacob and Alzheimer's diseases. Thus the PPI detection is important to investigate the intracellular signaling pathways, modeling of protein complex structures and gaining insights into various biochemical processes. Experimentally, PPIs are detected from a variety of techniques, including yeast two- hybrid systems, protein-fragment complementation assays (PCA), affinity purification/mass spectrometry, protein microarrays, fluorescence resonance energy transfer (FRET), and Microscale Thermophoresis (MST). Experimental technologies for PPI prediction are often laborious, time-consuming and costly. However, experimentally determined PPIs usually provide the basis for computational methods to predict interactions between two or more other protein molecules based on their homologous sequences across species. The computational method reduces the time and cost of PPI prediction drastically. Therefore, as an alternative, a computational approach for the identification of PPI is developed based on RF classifier. The accuracy of the Random Forest (RF) classifiers is 87.8% on test dataset which outperforms the other classifiers. Our proposed method achieves significantly better performances on both training and independent test dataset in Arabidopsis thaliana sequences. Thus the proposed method anticipated a very supportive computational resource for predicting PPI in Arabidopsis thaliana. Keywords: Protein-Protein Interaction, Random Forest, Classifier, Computational and Arabidopsis thaliana.

EC-066 Blockwise Robust Singular Value Decomposition for MRI Brain Image Denoising Md. Najibul Hasan1*, Md. Mamunar Rashid1 and Nishith Kumar1 1Department of Statistics, Bangabandhu Sheikh Mujibur Rahman Science and Technology University *Correspondence Email: [email protected]

Magnetic Resonance Imaging (MRI) is one of the high-dimensional high-throughput technologies that are playing a major role for diagnosing brain tumor. With the help of MRI images, brain tumor is diagnosed at advanced stages. Correct identification of brain tumor or abnormality, Image processing in MRI of brain is highly essential that can reduce the chance of fatal stage. Occasionally, these MR images are introduced with noise during acquisition which reduces the image quality and limits the accuracy in diagnosis. Therefore, preliminary diagnosis of MRI brain images from the hospital may not be always reliable for further analysis because of the presence of noise. Reduction/Elimination of noise in medical images is an important task in preprocessing which is one of the previous crucial parts for further image analysis. Although, several denoising techniques are available in the literature including median filter, wiener filter, wavelet filter, wavelet based wiener, non-local mean (NLM), wavelet based NLM, wavelet based weighted median filter etc. In this paper, we have measured the influence of the above methods for denoising. We have also proposed a new blockwise robust singular value decomposition technique for denoising image. The results of our analysis showed that blockwise robust singular value decomposition technique gives the better performance compared to the other methods. Therefore, our recommendation is to use the blockwise robust singular value decomposition technique for brain image denoising. Keywords: Magnetic Resonance Imaging (MRI), Preprocessing, Denoising, Image quality metrics, Wavelet Transformation.

133

EC-067 A study on parent’s perceived stress of unprivileged working and non-working children of special school. Nazmun Nahar Chowdhury, Farhana Hasan, and Md. Monimul Huq Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh Corresponding email: [email protected]

Background: Psychological health indicators are important for monitoring and evaluating the health status of communities. Perceived stress is a dynamic multidimensional concept, with a wide spectrum of causative and conducive factors. This study tries to analyze the relationship between parent’s stress level for their children according to their working status. This study was aimed to compare and verify the possible associations between the socio-demographic profile, stressor events and social support perceived by parents with working child, compared to parents with non-working child. Methodology: The present paper assessed and compared the type and level of perceived stress by the parents of 100 unprivileged working and non-working children of special schools across their social class. We use the Perceived Stress Scale questioner (PSS-Q) and questions was asked to the parents about their feelings and thoughts during the past month. . The severity and management of perceived stress were classified into five classes; never, almost never, sometimes, fairly often, very often. Descriptive statistics (Mean± SD) was utilized to find the level of perceived stress of parents of special school going unprivileged children. We use Chi square (χ2) test for the association of the variables. Statistical significance was accepted at p< 0.05. Results: The findings of− the present paper revealed that parents with non-working children was almost never (32%) in mental pressure; where parents of working children was very often (32%) in mental pressure with mean (3.36 ± 1.38) and (3.34± 1.56) respectively. Non-working children’s parents fairly often (40%; mean 3.20± 1.31) and working children’ parents almost never (36%; mean 2.98 ± 1.61) fails to make control important matter of their life. Both the parents was very often(52%; 78% and 72%; 74%) had the confidence to solve and tackle any problem with mean (3.96± 1.36; 4.66±0.75 and 4.42 ± 1.05; 4.46±1.01) respectively. Parents of non-working children never avoid (36%; mean 2.62 ± 1.56) any unwanted situation but the parents of working children very often tried to avoid(34%; mean 3.06±1.65) unwanted situation .Working children’s parents never (30%) angry with uncontrolled situation but their counterparts almost never (28%) angry with mean (2.84±1.56) and (3.14 ± 1.37) respectively. There was some significant association between the working status of unprivileged children and the perceived stress level of their parents. Conclusion: Stress is considered a risk factor for several chronic diseases such as hypertension, coronary artery disease, etc. We suggest that future studies might want to do a longitudinal analysis that follows a cohort of current child laborers over time compared with non laborers, and then assess whether their parents stress level about child labor continuation or termination hold when they become adults. Keywords: Stress level, working children, unprivileged, activity pattern.

EC-068 Analysis and Impact Evaluation of Machine Learning Algorithms for MRI Brain Image Classification Md.Mamunar Rashid1*, Nazibul Hasan1 and Nishith Kumar1 1Department of Statistics,Bangabandhu Sheikh Mujibur Rahman Science and Technology University, Gopalganj, Bangladesh. *Correspondence: [email protected]

Magnetic resonance imagining (MRI) is one of the sophisticated medical imaging techniques for diagnosing some particular diseases specially brain tumor, brain cancer etc. Sometimes, Cancer diagnosis is highly invasive, time consuming and expensive. However, one of the main advantages of MRI brain image is that it is a non-invasive and safer approach which saves time as well as cost for detecting brain tumor / abnormality. The traditional approach of brain tumor/ abnormality detection from MRI brain image is manual inspection, which may be inappropriate for large volume of images. Therefore, automated brain tumor / abnormality detection can solve the above problems. In the literature, there are several machine learning techniques including support vector machine, neural networks, k-nearest neighbor, naïve Bayes, linear discriminant analysis, random forests etc. for MRI brain image classification. However, the impact / influence of machine learning techniques for MRI image classification are different. Therefore, it is

134

important to find out the suitable machine learning technique for automated brain abnormality detection. In this paper, we have made a comparative study of machine learning algorithms for MRI brain image classification. This study has been based on three stages namely feature extension, dimensionality reduction and classification. We have also measured the performance of different machine learning techniques using classification accuracy, sensitivity and specificity etc. This analysis will help to choose the suitable machine learning technique for detecting brain abnormality automatically Keywords: Magnetic resonance imaging, Machine learning, Classification, feature extraction, dimension reduction.

EC-069 Level of depression and associated risk factors among the elderly in Sylhet district, Bangladesh Mohammad Kamal Hossain1, Md. Nazrul Islam2 and Mohammed Taj Uddin3 1Department of Statistics, Bangabandhu Sheikh Mujibur Rahman Science & Technology University, Gopalgonj, Bangladesh, Email: [email protected] 2Department of Statistics, Shahjalal University of Science and Technology, Sylhet, Bangladesh, Email: [email protected] 3Department of Statistics, Shahjalal University of Science and Technology, Sylhet, Bangladesh, Email: [email protected]

Objectives: Depression in the elderly is a public health problem. The objective of the study was to assess the prevalence of depression and to identify sociodemographic risk factors associated with the elderly population. Methods: Two stage cluster sampling for urban and multistage random sampling techniques for rural were employed, a cross‑sectional study of 944 elderly persons aged 55 or more were collected from the Sylhet district in Bangladesh. The tool used for data collection consisted of the Short Form of Geriatric Depression Scale (GDS-15). Results: The average ± SD age of the elderly is 62.66 ± 8.95 years. The prevalence of depression was 52.5% followed by mild or moderate depression 43.2%. Even 9.3% of the elderly were severely depressed. Findings of the study revealed that increased age, female, housewife or others, marital status, Muslim, nuclear type of family and living alone were found to be more depressed with their counterparts. Conclusions: The study shows that elderly having increased age, female, housewife or others, marital status, Muslim, nuclear type of family and living alone should be identified by screening of depressed with their counterparts. The study will be helpful for public health policymakers in Bangladesh. Keywords: Elderly, Depression, Geriatric Depression Scale (GDS), Bangladesh.

EC-070 Population Projection for Rajshahi City Corporation: A Comparative Analysis among the Projection Methods for Physical Development Plan Md. Awual Baksh1, Sharmin Akter2, Md. Ayub Ali3 1Deputy Urban Planner, TROYEE associates, Dhaka-1000, Bangladesh, [email protected] 2Graduate Student. Dept. of URP, RUET, Rajshahi, Bangladesh, [email protected] 3Professor, Dept. of Statistics, RU, Rajshahi, Bangladesh, [email protected]

In rapid urbanization era, Physical development plan (commonly used as Master Plan) is considered as all sorts of development guidelines for any urban area. Knowledge about current population and proper assumptions about future population are the base of almost all planning decisions. Population projection have a key role in the preliminary stages of preparing a plan, a proposal as to how land should be used as expansion and renewal proceed in the future. Considering this vast importance of proper population projection, the purpose of the study was to project and compare different population projection methods. Rajshahi City Corporation (RCC) was considered as a case study area and the study was conducted on the basis of secondary data from Bangladesh Bureau of Statistics (BBS). Mathematical trend extrapolation and the cohort-component model were applied for projection. For Analysis and projections, the study considered base population of RCC area 4, 49,756 and growth rate 1.25 at 2011 (BBS, 2011) and using the base year population made a projection of the study area population at five yearly intervals up to 2041. After applying different projection methods for the year up to 2041, the study found enormous difference of projected

135

population among the projection methods. In recommendation, one should be more careful to choose proper and suitable projection method in preparation of physical development plan as most of the development proposals depend on future population. Key Words: Population projection, Urban planning, Physical development plan, Mathematical trend extrapolation, Cohort-component model

EC-071 Generalized Vector Autoregression with Controlling Intervention and Volatility for Climatic Variables Md. Ashek Al Naim1 , Md. Abeed Hossain Chowdhury2, Md. Abdul khalek3, Md. Ayub Ali4 1 Research Student, Department of Statistics, University of Rajshahi, Email: [email protected] 2 Project Manager, BARC, Farmgate, Dhaka, Email: [email protected] 3 Associate Professor, Department of statistics, University of Rajshahi, Email: [email protected] 4 Professor, Department of statistics, University of Rajshahi, Rajshahi 6205 E-mail: [email protected] Corresponding E-mail: [email protected]

The purpose of the study was to build time series model and forecasting the climatic variables of Rajshahi district using the VAR model controlling intervention and volatility. There are 7 models for seven variables are found and stability was checked with proper modeling technique and validation. The fitted models were GVAR with GARCH(2,1) and intervention for Cloud coverage, GVAR with GARCH(3,1) and intervention for Humidity, ARIMA(3,0,0) for rainfall, GVAR with GARCH(2,1) and intervention for maximum Temperature, GVAR with ARCH(2) and intervention for minimum temperature, GVAR with intervention for sunshine, ARIMA(5,0,0) for wind speed. The stable models can be used to forecast the daily data which is beneficiary for people and policy makers. Government and policy makers can make people aware about the adverse effect of climate change. Keywords: VAR, ARCH, GARCH, ARIMA, GVAR, Stationarity, Stability, Climatic variables, Climate change

EC-072 Factors Influencing Healthcare Service in Khulna: A Case Study of Phultala Upazila Kamrul1, Md. Najmus1, Mumtahina1, Farzana1 and Md. Salauddin1 1Statistics Discipline, Khulna University, Khulna

The desired and right medical services for people are very important for a developed nation. Many private hospitals, as well as public hospitals, are operating in Bangladesh to provide proper health facilities. There is a contradicting result of satisfaction between the user of the public and private healthcare sectors. We investigate the influencing factors on the service of the healthcare system. A questionnaire was developed to identify the people’s participation in different types of hospitals according to their desired service, awareness for healthcare, their satisfaction level. We wanted to know the actual scenario of the general people in Khulna, so we have emphasized if there is an association between some variables from which we can conclude that the people are satisfied or not to the provided health services. Among the studied variables marital status, educational status, different types of hospitals used by respondents, waiting time for treatment, health personnel behavior are the significant factors that influence the healthcare service of the rural part. The findings revealed a complex interplay of factors influencing the rural and remote health facilities. Keyword: Countryside Area, Health Care Facilities, Affecting Factors.

136

EC-073 A Classification Approach to Predict Severity of Crime on Boston City Crime Data Sumaiya Tasnim1, Pritom Sarkar2, Ahammad Hossain3 and Md. Ayub Ali4 1B.Sc. (Hon’s) student, Dept. of CSE, Varendra University, Email: [email protected] 1B.Sc. (Hon’s) student, Dept. of CSE, Varendra University, Email: [email protected] 1Assistant Professor, Dept. of CSE, Varendra University, Email: [email protected] 2Professor, Dept. of Statistics, University of Rajshahi, Email: [email protected]

Crime is the execution of the thoughts of criminal mind. Which is why, in order to control crime, we need to chase the factors that have control over criminal mind. Till now there has been countless works on crime yet nowhere in near future it seems like researches on crime will be belittled at all. This fact only symbolizes our day-to-day struggle against crime and the heterogeneity of crime data. The purpose of this study was to predict the severity level of crime in Boston City, based on three basic components of time: season, weekday and 24-hours of a day. Time has been chosen as factor in our model as it has control over every human mind. The study was conducted by applying various classification algorithms on the secondary dataset of Boston city crime. Support Vector Machine (SVM), Decision Tree, Naïve Bayes and Binary Logistic Regression classifiers have been used here to predict the severity of crime. The outputs of these methods are captured by tenfold cross validation method and then compared to each other to find the one with best performance. This analysis portrays that Decision Tree has classified the highest number of correct classification by 72.91% followed by SVM with 70.88% and Binary Logistic Regression with 70.01% of correct predictions; whereas Naïve Bayes has least 69.10% of accuracy. This study has built a crime severity prediction model including a perceptible relative pattern among the basic time components with the objective to help law enforcement in dealing with the efficient analysis of the increasing volume of crime data which will ensure safety of people. Keywords: Crime Category, Feature Selection, Crime Classification, Crime Prediction

EC-074 Time Series Model Building and Forecasting on Maximum Temperature of Rajshahi District Kismot Ara1, Md. Ashek Al Naim2, Md. Abeed Hossain Chowdhuri3 Md. Abdul Khalek4, and Md. Ayub ali5 1 Research Student, Department of Statistics, University of Rajshahi, [email protected] 2 Research Student, Department of Statistics, University of Rajshahi, [email protected] 3 Project Manager, BARC, Farmgate, Dhaka [email protected] 4 Associate Professor, Department of Statistics, University of Rajshahi, Email: [email protected] 5 Professor, Department of Statistics, University of Rajshahi, Rajshahi 6205 E-mail: [email protected] Corresponding E-mail: [email protected]

The purpose of the study was to build time series model and forecast the pattern of the maximum temperature of Rajshahi district. The model building process was ARMA. Appropriate model building technique with validation and stability of the model were checked and ensured repeatedly. SARMA (3,2)

(2,0)12 was the final model for forecasting. It was found that the fitted model was significant and stable for future. By forecasting with this model, the maximum temperature will be increased by 1 degree within 2022. The forecasting result is alarming for the mankind of this area. Based on this forecast, Government should take proper steps to concern people about the bad effect of increasing temperature and come forward to save the biodiversity. Keywords: ARMA, SARMA, Maximum temperature, Stationarity, Model validation and Model stability

EC-075 Effect of socio-economic and anthropometric factors on HIP among children in Rajshahi district: A cross-sectional study Nishat Zahan1, Ismail Hossain1, Md. Nurul Islam1, Md. Ripter Hossain1 Abu Sayed Md. Al Mamun1, and Md. Golam Hossain1 1Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi

Nutrition, genes and environment are the three major factors that impact a child’s development. Nutritional status of children is an important indicator of their health status of a particular population. Globally, it contributes to 45% of deaths of children aged (6-10 years) children, and most of the death occurs in

137

developing countries. Study on nutritional status of children is important for determining risk factors and current prevalence of malnutrition in a particular country to take initiative for remove or reduce malnutrition among children. The term malnutrition` (overweight or obese) has a great importance in public health community. Malnutrition has been shown to be a risk factor for disease. Obese children have been shown to be risk factors for high blood pressure and high cholesterol, which are risk for cardiovascular disease (CVD). Obese children are more likely to become obese adults. The cross-sectional study sample consisted of 1200 (boys 686 and girls 514) at the primary school, kintegartain, English medium and madrasah of Rajshahi, Bangladesh. Sample size was determined using an appropriate statistical formula. Students were interviewed from May to October 2015 using a semi-structured questionnaire containing pre-coded and open-ended questions. Height and weight of the children as well as their parents has been measured. Chi-square test and binary logistic regression analysis were used in this study. The data was analyzed with Statistical Package for Social Sciences (SPSS) version 20.0. The mean age of the student was 7.77±1.40 year with range 6 to 10 years. The mean HIP of school going children was 58.83±8.96 cm. A value of p<0.05 was regarded as statistically significant in the analysis. HIP circumference of school going children (age, 6-10 years) was classified into two classes according their nutritional status such as (i) under nutrition (HIP<60 cm) and (ii) normal (HIP≥60 cm). More than 65% students were under nourished while 34.20% were normal or healthy. The binary logistic regression model demonstrated that higher educated mother’s children more likely to get normal (HIP ≥60 cm) than secondary educated mother’s (AOR=0.948, 95% CI: 0.358-2.507; p>0.05) children. These results suggest that there are dual burden of malnutrition among primary school children in Rajshahi district, Bangladesh. Under-nutrition is the common phenomenon in Bangladeshi children especially for girls, living in poor family; the other hands over-nutrition is common for boys and living in rich family. Consequently, malnutrition can be considered as one of the major health problems of Bangladeshi children and requires attention. Keywords: Socio-economic factor, Chi-square test, Logistic regression, Rajshahi Submission for Student Poster Competition? Yes

EC-076 Time Series Model Building and Forecasting Population of Bangladesh Md. Ashek Al Naim1 and Md. Ayub Ali2 1 Research Student, Department of Statistics, University of Rajshahi, Email: [email protected] 2 Professor, Department of statistics, University of Rajshahi, Rajshahi 6205 E-mail: [email protected] Corresponding E-mail: [email protected] The purpose of the study was to model building and forecast the population of Bangladesh through the model building process using ARIMA with GARCH. Proper modeling technique with validation and stability of the model were checked and ensured respectively. ARIMA (6,0,0) with covariates (year) was the final model for forecasting. It was found that the fitted model was significant and stable for the future. Using this model the population of Bangladesh will be (19,94,40,470 ± 10,000) by 2030. On the basis of this forecast government may take necessary steps for achieving SDG’s. Keywords: Population, ARMA, GARCH, SDG, Model stability and validation, Non-stationarity

EC-077 Study on Menstrual disturbances among menarcheal girls (age, 10-15 years) in Rajshahi district, Bangladesh Dolly Khatun1, Rajwana Parvin1, Abu Sayed Md. Al Mamun1 and Md. Golam Hossain1 1Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi

Menstrual disturbances generally affect the daily activities of young females. The aim of this study was to determine the effects of anthropometric measures and socio-demographic factors on menstrual disturbances among school going students (age, 10-15) in Rajshahi district, Bangladesh. A total number of 628 menarcheal girls were considered as a sample in this study for investigation their menstrual disturbances. Two menstrual disturbances of school girls (i) irregular menstrual cycle and (ii) painful menstrual cycle were investigated in this study. Among the menarcheal girls, 23.2% were suffered from irregular menstruation cycles and 77.5% from menstrual pain. Girls who got menarche earlier they were more likely to suffer irregular menstruation (p<0.05). Higher educated father’s daughters suffered more by irregular menstruation cycles (p<0.05). Girls of very high bleeding faced menarche earlier (p<0.01) and whose duration of staying menstruation is high

138

they were suffered from irregular menstruation cycles (p<0.05). Girls of getting early menstruation suffered menstrual pain significantly (p<0.05) more than their counterparts and girls of urban area were more likely to get menstrual pain than their counterparts (p<0.01). Girls whose mode of delivery were caesarean they were more likely to get menstrual pain (p<0.05). Higher educated father’s daughters were more likely to have pain during menstruation (p<0.01). Girls of very high bleeding got more menstrual pain (p<0.01) and whose duration of staying menstruation is high they were suffered from menstrual pain than their opposite parts (p<0.01). Girls having higher BMI suffered menstrual pain more than other girls (p<0.05). Girls having less number of siblings were more likely to have menstrual pain (p<0.05). Service holder fathers daughters (p<0.01) suffered pain more and girls of higher income were more likely to suffer menstrual pain (p<0.01). Finally stepwise logistic regression (forward elimination) demonstrated that residence was the most significant factor for early age at menarche and how much bleeding during menstruation was the most significant factor for irregular menstrual cycles and menstrual pain both. Keyword: Menstruation, Girls, Logistic Regression, Rajshahi District. Submission for Student Poster Competition? Yes

EC-078 Body mass Index of University students and gender differential: survey in Rajshahi University, Bangladesh Jannat Ara Nipa, Md. Kamrul Hasan, Abu Sayed Md. Al Mamun and Md. Golam Hossain 1Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh

Body mass index (BMI) can be used as an indicator for health status of a population. BMI is particularly important for developing countries like Bangladesh where health is a major public issue. The aim of the study was to investigate the BMI among university students in Bangladesh and gender differential. Data were collected from 911 (male 727 and female 184) Rajshahi University students from January to June, 2014 using stratified sampling technique. Students t-test and z- proportional test were used in this study. In this study male and female students were analyzed separately. The results suggest that more than 26% female students have been suffering from chronic energy deficiency, and under nutrition is the common phenomenon in Bangladesh especially for female population. Consequently, under nutrition can be considered as one of the major health problems of Bangladeshi university students and requires attention. Keywords: Nutritional status, University students, Gender. Submission for Student Poster Competition? Yes

EC-079 Prevalence and risk factors of hypertension among Rajshahi university students: a cross- sectional study Moktadir Billah Mutasim, Lamia Tabassum Abu Sayed Md. Al Mamun, and Md. Golam Hossain 1Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh Early detection and treatment of hypertension can help to reduce the morbidity and mortality rate due to heart disease and stroke in a population. The purpose of this study was to determine the prevalence and factors associated with hypertension among university students in Bangladesh. Data were collected from 911 Rajshahi University students using a stratified sampling approach between January and June, 2014. The Chi- square test and multiple logistic regression analysis were utilized in this study. The prevalence of pre- hypertension and hypertension among university students in Bangladesh were 42.50% (systolic 46.00%% and diastolic 37.10%) and 25.60% (systolic 9.44% and diastolic 23.93%) respectively. The Chi-square test showed that more than 50% of the selected variables were significantly associated with hypertension. Multiple logistic regression model showed that the male students were more likely to display hypertension than female students (p<0.01). Overweight or obese students were also more likely to display hypertension when compared to normal (p<0.01) and underweight students (p<0.01). The results in this study suggest that more than one fourth university students have been suffering from hypertension. Gender and body mass index are two important predictors for getting hypertension. Awareness about a healthy life style may reduce the rate of hypertension among university students. Keywords: Hypertension, University Students and Multiple Logistic Regression. Submission for Student Poster Competition? Yes

139

EC-080 Robust QTL Analysis Based on Robust Estimation of Bivariate Normal Distribution Md. Jahangir Alam1*, Md. Ripter Hossain1, S.M. Shahinul Islam2 and Md. Nurul Haque Mollah1* 1Bioinformatics Laboratory, Department of Statistics, University of Rajshahi, Rajshahi 6205, Bangladesh, 2Institute of Biological Science, University of Rajshahi, Rajshahi 6205, Bangladesh *Corresponding authors: Md. Jahangir Alam ([email protected]), [email protected] Simple interval mapping (SIM) is one of the most popular approaches for genome wide single quantitative trait locus (QTL) analysis. Maximum likelihood (ML) and ordinary least squares (OLS) regression are widely used methods for SIM. However, these approaches are very sensitive to outliers and provide misleading results in presence of phenotypic contamination. For backcross and F2 population, there exist robust approaches of SIM which is based on the estimation by maximizing the -likelihood function of the model parameters. We have introduced a new robust approach of SIM for single-trait QTL analysis which is based on the assumption that the phenotype and conditional probability of QTL genotype given the marker genotypes have jointly bivariate normal distribution. In our proposed approach, we estimate the model parameters based on the properties of bivariate normal distribution and using the ML estimates of parameters of the distribution. We robustify the parameter by maximizing -likelihood function of the parameters of bivariate normal distribution. Our proposed method can identify the same QTL positions as identified by the existing robust methods of SIM. In the proposed method, robust LOD statistic has been calculated using robust estimates of residual variances which are based on robust estimates of distribution’s parameters. Keywords: Simple interval mapping, maximum likelihood, -likelihood function, bivariate normal distribution, LOD statistic.

EC-081 Survey on knowledge and practice regarding antibiotic use among rural peoples in Rajshahi district, Bangladesh: A cross sectional study Dipongcor Sarkar Dipon1, Md. Jakirul Islam1, Md. Nurul Islam1, Md. Ripter Hossain1, Abu Sayed Md. Al Mamun, and Md.Golam Hossain 1Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh

Decreasing antibiotic effectiveness is a big problem in Bangladesh. The aim of the study was to assess the knowledge and practice about antibiotic among adults in rural areas of Rajshahi district in Bangladesh. A cross-sectional study was performed on 600 rural adult peoples. Data were collected in December 2015 from the different rural areas in Rajshahi district, using a standard semi-structured questionnaire. Multi stages random sampling was used for selecting our sample. Chi-square test was utilized to find the factors which were significantly associated with peoples’ knowledge and practice on antibiotic use. In this study male and female participants were 49.33% and 50.67% respectively. The results revealed that only 6.67% people were aware of antibiotic, and 90.5% community people did not have knowledge about the proper situation for using of antibiotic. About 70% participants did not have proper knowledge about the completion of antibiotic course. A few numbers (0.50%) of participant had properly followed health service provider's suggestion about using of antibiotic. Chi-square test demonstrated that gender, age, residence, monthly family income, and education were significant factors for the knowledge and practices on antibiotic use. The knowledge level about proper use of antibiotic was found very poor among rural peoples in Rajshahi district, Bangladesh. We noted that some modifiable factors are associated with the lack of proper knowledge and practices on antibiotic use. Specific intervention is required to improve basic knowledge on antibiotic use among Bangladeshi population. Keywords: Awareness, Antibiotic, Rural people and Bangladesh. Submission for Student Poster Competition? Yes

EC-082 The Use Of Polygenic Risk Scores As An Instrumental Variable In Mendelian Randomization Studies: A Systematic Review Shabab Noor Islama and Md. Jamal Uddina a Department of Statistics, Shahjalal University of Science & Technology, Sylhet-3100, Bangladesh.

Polygenic risk scores (PRS) has increasingly been used as an instrumental variable (IV) in Mendelian randomization studies to control for unmeasured confounding or measurement errors. However, in empirical

140

studies, the usefulness of the PRS as an IV and/or proper reporting of the methodological issues (e.g. assessing assumptions) are sparse. Thus, our objective was to review published clinical papers and identify reporting problems and usefulness of the PRS and then provide recommendations for clinicians. We conducted a systematic review using the clinical articles published between 2009 and 2019. We searched mainly in PubMed databases. We extracted the information from each study reported about instrumental variables including particular tests used to check assumptions, modeling approach used for estimation, etc. Using our inclusion/exclusion criteria, in our pilot review, we identified 63 studies. In our review, we noticed a tendency that most of the studies (71.4%) reported assumptions number 1 but only 37 (58.7%) studies reported the appropriate tests (e.g., R2, F-test) to check the validity of the assumptions. 47.6% clearly stated and discussed the 2nd and 3rd assumptions but among those only 50% of studies clearly mention MR-Egger for checking the 3rd assumption but no studies refer any tools for checking 2nd assumption, the only logical explanation was provided. For the parameter to be estimated from IV models, only 23% reported the odds ratio. Moreover, 33.3% of studies used a two-stage least square method for estimating instrumental variables. In addition, 41.3 % of studies conducted sensitivity analysis to illuminate the robustness of estimates for the violations of the untestable conditions. In conclusion, we find that incompleteness of the justification of the assumptions of the instrumental variable is a common problem in our selected studies. This may misdirect the objective of the study in the wrong way. So, we point out that fundamental issue in an IV method is not the decision of technique but instead the selection of IVs and the evaluation of the IV assumptions. Keyword: Instrumental variable, Mendelian Randomization, Assumption Submission for Student Poster Competition: Yes

EC-083 Lack of general knowledge and its associated factors of infectious diseases among adults in rural area of Rajshahi district, Bangladesh: A cross sectional study Md. Alamgir Hossen1, Md. Jahangir Alam1, Abu Sayed Md. Al Mamun1 and Md. Golam Hossain1 1Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh

Infectious disease remains a problem for many less developed countries. The aim of the study was to assess the knowledge on preventive measures on infectious diseases among adults in the rural areas in Rajshahi, Bangladesh. A total number of 1,161 samples were considered in this study. Data were collected from rural areas of Rajshahi district using semi-structural questionnaire. Logistic model was utilized in this study. We observed that more than 79% participants had some knowledge on how to prevent diarrhea, but knowledge on prevention was low for worm infestation (40.6%), skin infection (39%), eye infection (11.2%) and measles (26.1%), and very low for filariasis (2.4%) and kala-azar (1.8%). Multiple logistic model demonstrated that younger participants was more likely to have preventive knowledge about diarrhea (p<0.05) than older, but the opposite result was found regarding worm infestation (p<0.05). Female had more aware of preventive knowledge on diarrhea disease and worm infestation than male (p<0.05). Higher educated respondents were more likely to get preventive knowledge about infectious diseases than uneducated persons (p<0.05). Our study shows that the level of preventive knowledge about infectious disease is very low among the adults in the rural areas of Bangladesh and requires attention. Keywords: Awareness, Infectious diseases, Rural area and Bangladesh. Submission for Student Poster Competition? Yes

EC-084 Factors influencing cigarette smoking behavior among university male students in Bangladesh: Survey in Rajshahi University Md. Nurul Abshar1, Md. Nurul Islam1, Md. Ripter Hossain 1, Abu Sayed Md. Al Mamun Md. Golam Hossain 1Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh

Cigarette smoking is a risk factor for various diseases. This epidemiological study was designed to determine the prevalence and risk factors of cigarette smoking behavior of university male students in Bangladesh. A total of 600 students were selected from Rajshahi University during the period of October to December, 2016, using multistage random sampling. Chi-square (χ2-test) and logistic regression model were

141

utilized in this study. The present study revealed that the prevalence of cigarette smoking behavior among Rajshahi university students was 9.50%. The χ2-test demonstrated that monthly family income (p<0.05), fathers’ smoking status (p<0.01), brother(s)/uncles’ smoking status (p<0.01), type of family (p<0.01) and family residence (p<0.01) were significantly associated with students’ cigarette smoking behavior. Logistic regression model showed that students had higher chance to become cigarette smoker who came from rich family (p<0.05) and urban areas (p<0.01). It was found that smoker fathers’ son (student) was more likely to be smoker than their counterparts (p<0.01). A student had more chance to be smoker if his brother(s)/uncle(s) was also smoker (p<0.01). Some modifiable factors of students’ cigarette smoking behavior were found in this study. These factors might be considered for removing cigarette smoke among university students in Bangladesh. Keywords: Socio-economic factor, smoking behavior, university students and Bangladesh. Submission for Student Poster Competition? Yes

EC-085 Anthropometric measures influencing facial shape of adult Japanese male student: Multiple regression analysis Md. Moazzem Hossain Bappy1, Md. Refat Ahsan Rabby1, Md. Imran Hossain1 Md. Nurul Islam1, Md. Ripter Hossain1, Abu Sayed Md. Al Mamun1, Md. Golam Hossain1, 1Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh The shape of human head and face depends on various craniofacial measurements. The aim of the study was to investigate the effect of anthropometric measures on facial shape of Japanese adult male students. In this study, prosopic index (PI) was used as a measurement of facial shape. The relationship between PI and several anthropometric measures were examined by multiple linear regression model. Stepwise linear regression analysis was used to determine the most important predictors for PI. The majority (26.19%) of the Japanese adult male facial shapes were leptoprosopic (long face) and mean prosopic index was 84.19. The results of regression analysis suggest that the bigonial breadth (p<0.01), head length (p<0.01), head height (p<0.01), head circumference (p<0.01), stature (p<0.01), head breadth (p<0.01), and minimum frontal breadth (p<0.01) were the most important predictors for facial shape of Japanese adult male students. Keywords: Anthropometric measures, facial index and Japanese students. Submission for Student Poster Competition? Yes

EC-086 Effect of nutritional status on academic performance of high school students (level 6 – 8) in Rajshahi district, Bangladesh: A cross sectional study 1Mahbub Elahi1, Md. Imran Hossain1, Abu Sayed Md. Al Mamun1 and Md. Golam Hossain Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh School education is a vital stage in the development of the consciousness and personality of children. Nutritional status is an indicator of health status in a particular nation. The impact of nutritional status on academic performance among high school students (level 6 to 8) are poorly documented in Bangladeshi. The objective of this study was to investigate the effect of nutritional status on academic performance of level 6 to 8 students in Rajshahi district, Bangladesh. Data was collected from the differences schools in urban and rural setting in Rajshahi district, Bangladesh. Multistage sampling was used for selecting the sample. A total of 405 students (male, 195 and female 210) were participated with their mean age 13.11±1.15 years (age range, 10 to 17 years) (male 13.16±1.16 years and female 13.06 ±1.09 years). This study raveled that the prevalence of under nourished and over nourished were 63% and 3.20%, male was more under nourished than female and rural students was more under nourished than urban. The average grade point average (GPA) of 6 to 8 level students was 3.70±0.80, and it was noted that the average GPA of urban students was higher than rural students (p<0.01). It was found the positive relationship between BMI and academic results of students. We observed that there was an increasing tendency of academic results among students with increasing their nutritional status. We found that a remarkable number of high school students had been suffering from chronic energy deficiency in Rajshahi district, Bangladesh. It was also observed that healthy students obtained good academic results than unhealthy students. This study may suggest health authorities of Bangladesh for special looking our school going students’ health. Keywords: Nutritional status, academic performance, school student and Bangladesh. Submission for Student Poster Competition? Yes

142

EC-087 Nutritional status of Bangladeshi adults: Gender differentiate 1Md.Bayazid Hossen1, Md. Alhaz Uddin1, Md. Moidul Islam1 Waris Khan1, Abu Sayed Md. Al Mamun1, Md. Golam Hossain1 1Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh

Malnutrition is the important key risk factors for the various diseases. In this study, body mass index (BMI) was used to measure nutritional status for adults. The aim of the study was to differentiate nutritional status between male and female in Bangladesh. Data was extracted from Bangladesh Demographic and Health Survey (BDHS-2011) dataset. Independent sample t-test, Z-proportional test, chi-square (χ2) test and multinomial logistic regression were used in this study. This study revealed that the prevalence of under and over nutrition among Bangladeshi adults aged 16-34 years were 23.6% and 14.3% respectively. It was found that the prevalence of under nutrition among men (26.3%) was significantly (p<0.01) higher than that of women (23.3%), while the prevalence of over nutrition among women (15.2%) was higher than that of men (7.2%). Height and weight of adult male was significantly (p<0.01) higher than female, but the mean value of BMI of female (21.32±3.59 kg/m2) was significantly higher (p<0.01) than male (20.54±2.83 kg/m2). Urban male and female were higher BMI than rural male and female (p<0.01). Multinomial logistic regressions exhibited that living location (division), education level, toilet facilities, marital status and wealth index were risk factors for malnutrition for both sexes. This study demonstrated that a remarkable number of both male and female were malnourished in Bangladesh. This study suggested that controlling some modifiable factors such as household wealth quintile, education, working status and residence might reduce the prevalence of malnutrition among adults in Bangladesh. Keywords: Gender differentiates, Nutritional Status, Adults, Bangladesh, Logistic regression Submission for Student Poster Competition? Yes

EC-088 Study on infertility and related factors among women in Bangladesh: Survey in Rajshahi City Meherun Nesa1, Shah H.A. Mahdi1, Fatema Siddiqua2, Sabiha Sultana1, Abu Sayed Md. Al Mamun3 and Md. Golam Hossain3 1Department of Zoology, University of Rajshahi, Rajshahi 6205, Bangladesh 2Islami Bank Medical College, Rajshahi 6203, Bangladesh 3Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh

Infertility is a common medical problem among women, and in many traditional societies, women are usually considered to be responsible if married couples are unable to bear children. The objective of the present study was to investigate the problem of infertility among women seeking treatment at Rajshahi City in Bangladesh. A total number of 500 infertile women were considered in the present study. Data were collected from Motherland Hospital and Infertility Centre at Rajshahi in Bangladesh during the period of February to August 2016. In this study, convenient sampling was used for selecting sample, t-test and z- proportion test were utilized in this study. In the current study revealed that more than 60% women had been suffering from the primary infertility. More than 68% infertile women were trying for getting baby for last 5 years. Most of the women (82.8%) did not attempt to get pregnant at specific period of their menstrual cycle. Many of them (79.6%) were not aware about the concept of fertile period. Larger percentage of the individuals was educated housewives. It was noted that husbands’ age was significantly larger than infertile women age (p<0.01). Authorities should provide basic education on fertile period to educated housewives, and organize more general campaign on infertility for working mothers and those who were less educated. Keywords: Infertility, Women, Socio-demographic factors, Rajshahi City, Bangladesh

143

EC-089 Reproductive health care services for women at community clinic in Rajshahi city: A cross sectional study Momanin Mohammad Saqlain1 , Sheikh Moin Uddin2, Md. Masud Rana3 Abu Sayed Md. Al Mamun4 and Md. Golam Hossain4 1Rajshahi TB and Leprosy Control Project (RTLCP), Damien Foundation Bangladesh, Email: [email protected] 2Faridpur TB and Leprosy Control Project (FTLCP), Damien Foundation Bangladesh, Email: moin.lipu@ gmail.com 3Department of Population Science and Human Resource Development, University of Rajshahi, Rajshahi 6205, Bangladesh, Email. masud [email protected] 4 Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh, E-mail: [email protected]

Abstract: In Bangladesh has achieved remarkable successes in improving the health of the population, some health indicators remain poor. One factor contributing to this situation is low use of primary and community health services. The aim of the study was to assess the general knowledge community clinic and its reproductive health services. A cross sectional descriptive study was carried out for a view to assess knowledge about reproductive health care services for women at community clinic in Rajshahi. The sample size was 307. Study showed that (53.7%) were in the age group of 21-30 years, monthly family income (87.6%) was taka 6001-10000, education level 54.1% was class VI-XII, among them (79.2%) were housewife, 99.0% of the respondents heard about CC, (97.1%) of the respondents received services from the CHCP. It was found that majority (98.7%) of the respondents knew about the health care services provided by CC, majority (99.3%) of the respondents knew about the services provided for MCH, majority (99.3%) of the respondents knew about services provided for family planning by CC, majority (95.8%) of the respondents knew about the services provided for health education, majority (97.7%) did not know about the services provided for EOC, majority (83.1%) of the respondents knew about services provided for safe motherhood, majority (90.6%) of the respondents knew about the services provided for referral to UHC. The relationship between educational status of the respondents and going to CC, types of care was found statistically non-significant (p>0.01). The relationship between occupation of the respondents and by whom the service was provided, type of care was found statistically significant (p<0.01). It was found that 98.70% of the respondents were Muslim, 94.46% of the respondents were married and 52.44% of the respondents were from joint family. This study provided some important information which might help the concerned authority to take appropriate measures to improve the health status of the women in Rajshahi district. Keywords: Awareness, Reproductive health, community health services and Rajshahi City.

EC-090 Factors Associated with Control of Diabetes among Type-2 Diabetes Patients in Bangladesh Md. Abu Sayem1, Md. Rocky Khan Chowdhury2, Abu Sayed Md. Al Mamun3 and Md. Golam Hossain3 1 National Tuberculosis Control Program (NTP), Directorate General of Health Service (DGHS), Bangladesh. Divisional TB Expert – Khulna Division, National Tuberculosis Control Program (NTP), Bangladesh, E-mail: [email protected] 2 Department of Population Science and Human Resource Development, University of Rajshahi, Rajshahi, Bangladesh. E-mail: [email protected] 3Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh [email protected] and [email protected]

The control of blood glucose is associated with several socio-demographic, behavioral and biomedical factors. The objective of the study was to identify the factors influencing the control of diabetes. A cross- sectional study was performed among 300 diabetic patients using purposive sampling methodology in Diabetic hospital, Rajshahi, Bangladesh. Chi-square test and logistic regression analysis were used to find out the associations of diabetes control with selected factors. Our results show that around 82.0% patients could control diabetes. According to chi-square test, education, gender, monthly family income, regularity of drug intake, regular check-up, vertigo and high blood pressure (HBP) were found as associated with diabetes control. Multivariable logistic regression model suggested that gender, duration of diabetes, regular drug intake, regular check-up, physical activity, vertigo and HBP have significant effect on diabetes control. As

144

diabetes is a major public concerning issues in Bangladesh, Government should be given top priority in its health agenda. Key words: Diabetes control, Socio-demographic, Logistic regression

EC-091 Computational Prediction of PPI Mapping on Homo sapiens under the Amino Acid Composition Encoding Scheme Md. Merajul Islam, Samme Amena Tasmia, Md. Jahangir Alam, Fee Faysal Ahmed, Md. Hadiul Kabir and Md. Nurul Haque Mollah* Bioinformatics Lab., Department of Statistics, Rajshahi University, Rajshahi-6205, Bangladesh; *Correspondence: Md. Nurul Haque Mollah ([email protected])

Background and Objective: Protein-protein interactions (PPIs) handle a wide range of biological processes, including cell-to-cell interactions and metabolic and developmental control and plays key role in predicting the protein function of target protein and drug ability of molecules. To understand the molecular structure and function of human, the identification of protein-protein interaction is important. However, the performance of other existing methods for identification of PPIs is not satisfactory level. Thus, it is highly urgent to develop computational methods that can efficiently and accurately predict the potential PPIs based on human protein sequence information. Methods: In this study, firstly, amino acids composition (AAC) encoding strategy is used to convert Human protein sequences to numeric data. After generating encoded protein sequence features, we selected vital features using Wilcoxon rank sum feature selection method. Then Naïve bayes (NB), AdaBoost (AB), Logistic regression (LR), Random Forest (RF), Support vector machine (SVM), K-Nearest Neighborhood (K-NN) and Neural Network (NN) classifiers were trained with the selected vital features. Main Finding: Among seven classifiers, RF classifier provides most accurate predictions of PPIs and obtained the average performance scores of ACC, SN, SP, PPV, NPV, MCC and AUC are 92.0%, 97.0%, 82.0%, 91.5%, 93.1%, 83.6% and 0.951 respectively under the 10-fold cross validation. Conclusions: The proposed method anticipated a very supportive computational resource for predicting PPIs in Human species and can be taken as a supplementary tool for future proteomics research to contribute the development of new Drug in the health sectors. Keywords: Protein-Protein interaction (PPI), Amino Acid Composition, Feature Selection, Random Forest and PPI prediction.

EC-092 Weather Forecast based Rice Crop Management System in Bangladesh: An Experimental Approach for T. Aus Season Niaz Md. Farhat Rahman, Md. Abdullah Aziz, Md. Abdullah Al Mamun, Afroza Chowdhury, Rokib Ahmed, Md. Mofazzel Hossain, ABM Zahid Hossain, Mohammad Ashik Iqbal Khan, Md. Khairul Alam Bhuiyan, Md. Mozammel Haque, Tuhin Halder, Md. Ismail Hossain and Md. Shahjahan Kabir Agrometeorology and Crop Modeling Laboratory Bangladesh Rice Research Institute (BRRI), Gazipur-1701, Bangladesh

Strengthening the early warning system to forecast extreme weather and provide action-oriented advisories is an essential step to increase rice yield as well as enhancing the farmer’s income. With this context, the study was conducted to assess the performance of weather forecast based rice crop management practices in comparison to traditional management. In this study, four districts during T. Aus season (April to August) were selected based on their diverse agro-climatic conditions: Rajshahi (cold & heat), Barishal (coastal flooding & waterlogging), Rangpur (cold and submergence) and Gazipur (favorable condition). Most popular and high yield potential varieties BR26, BRRI dhan48, and BRRI dhan65 were used in this regard. Randomized Complete Block (RCB) design with three replications was followed as field experimental design. Also, grain yield data was adjusted to 14% moisture content. The weather research and forecasting (WRF) model was used to generate weekly weather forecasts on rainfall (mm), temperature (0C), humidity (%) and wind speed (m/s). Based on weather forecasting advisories were made and practiced into the respective experimental field of all locations through email communication. Linear mixed model (LMM), mean comparison test was performed for four locations to estimate the mean yield, mean square error, and

145

coefficient of variation and find out the overall and location-specific performance of forecast-based management practices. Results revealed that throughout the T. Aus season weather conditions for all locations were more or less similar pattern. Rajshahi had the most extreme weather condition, always high temperatures and low rainfall and humidity. High rainfall was found in the Gazipur district. There is a good relationship between temperature and rainfall i.e. vice-versa. July is the highest rainfall and low-temperature month. On the other hand, May is the highest temperature and low rainfall month. LMM analysis showed there was a significant difference in management practices and the interaction of location and management practices meaning that yield of rice varieties varies in case of different management practices and yield also varies in different locations under different management practices. From the mean separation test, weather forecast based rice crop management practices performing better which produces 4.75 t/ha than traditional produce 3.91 t/ha. So, the research demonstrated high potential for field implications of the forecast based rice management practices resulted in 21.48% higher rice yield than the control plots with traditional management practices. Hence, analyses suggested weather forecast based rice crop management practices may be recommended for all locations in T. Aus rice cultivation having a similar environment as that of the experimental site. Keywords: Combined ANOVA, LMM, rice advisory, rice varieties, and weather forecast.

EC-093 Determinants of Academic Performance on Secondary School: A Study in Rajshahi City K. M. Barkullah1, Md. Rafiqul Islam1, Md. Masud Rana2, Abu Sayed Md. Al Mamun2 1Dept. of Population Science and Human Resource Development, University of Rajshahi-6205, Bangladesh 2Dept. of Statistics, University of Rajshahi-6205, Bangladesh

Abstract: In this era of globalization and technological revolution, education is considered as a first step for every human activity. It plays a vital role in the development of human capital and is linked with an individual’s well-being and opportunities for better living. Socio economic and demographic factors influenced the academic performance of the students. The aim of this study was to find out the influencing factors affecting academic performance of secondary school students. A cross-sectional survey was performed on 200 (114 male and 86 female) college going students. Data were collected by using a semi structured questionnaire about different factors relating to academic performance of students. The purposive sampling technique was used to select the students for this study. The academic performance was gauged by the result of their secondary school certificate (S.S.C) examination. Descriptive, chi-square test and binary logistic regression model were applied to investigate the effects of different factors on students’ achievement. Among 200 school students, male and female were 57 % and 43% respectively, and 49.0% students were got GPA 5. Higher educated and service holder father’s son and daughter were got GPA 5, 57.2% and 67.1% accordingly. Most of the service holder’s mother son and daughter got GPA 5, 79.4 % respectively. Services holder mothers son and daughter (AOR=5.159; 95% CI: 1.003-26.540, p< 0.05) times more likely to have GPA 5. Chi-square -test demonstrated that father, mother education, father and mothers occupation, monthly family income, GPA got in JSC, study section and expected results in SSC were associated with academic performance of school students. The present study demonstrated that father and mother education can play important role in their son and daughter academic performance of school students. Keywords: Academic performance, Chi-square test, Binary logistic regression model

EC-094 Projected Yield Potential of Rice Varieties towards Increasing the Productivity in Bangladesh 1Niaz Md. Farhat Rahman, 2Md. Ahadur Rahman, 1Mohammad Chhiddikur Rahman and 1Md. Shahjahan Kabir 1Bangladesh Rice Research Institute, Gazipur, Bangladesh 2Shahjalal University of Science and Technology, Sylhet, Bangladesh

For keeping the pace of rice production and future food security, the arrangement of high yielding varietal release will be the key initiative. Considering this context the objective of this study is to estimate expected actual farm yield (t/ha) and projected variety potential yield (t/ha) of rice varieties up to 2030 and the required expected rate of increment (%) per year to achieve the targets. For projecting varietal yield potential

146

from 2020 to 2030, 2018 was considered as the base year, area and production of that year were collected from BBS, 2018 as well as projected area and production data were collected from ‘Rice Vision 2050 and Beyond’. Potential yield of BRRI released varieties was collected from the modern rice cultivation of BRRI and the average potential yield was calculated by arithmetic mean using rice varieties of all the seasons. In order to calculate the rate of increment the log-lin model and in case of projection compound exponential growth model were employed in the analyses. The estimation was made for the projected year 2020, 2025 and 2030. Analytical results revealed that in order to reach the target to feed the projected population the yield of Aus, rice varieties should be increased by 215, 110 and 93 kg/ha at 2020, 2025 and 2030 respectively. It should mention that the Aman rice varieties are performing beyond the projected yield of 2020 and 2025. Only 1% increase in the yield will be required to meet the upcoming requirement during 2030. The rice breeders have to overcome serious challenge to develop Boro rice varieties with much higher yield potentiality. On average, we have to develop Boro rice varieties with an increment of the yield at the rate of 550, 233 and 181 kg/ha by 2020, 2025 and 2030 respectively. So, we can conclude that the high yield potential variety release can ensure the required production up to 2030. Keywords: Mathematical model, rice area and production, rice productivity and Rice Vision 2050 and Beyond.

EC-095 Determining Comorbidity Under-five Children in Bangladesh: A Multilevel Modeling Approach Ruhul Amin1, Md. Atiqul Islam2 1Department of Statistics, Shahjalal University of Science and Technology, Sylhet-3114, Bangladesh, Email: [email protected] 2Department of Statistics, Shahjalal University of Science and Technology, Sylhet-3114, Bangladesh, Email: [email protected]

Background: Challenges of health care systems, comorbidity impact, and prevalence are very high in under- 5 children globally. Therefore, the aim of study is to identify the risk factors of comorbidity (fever, diarrhea, and ARI-acute respiratory infection) of under-5 children in Bangladesh taking into account the clustering structure. Methods: Data were extracted from the Bangladesh Health and Demographic Survey 2014. Generalized Linear Mixed Models (GLMMs) were applied to investigate the risk factors of child comorbidity. Result: The study observed that the prevalence of fever, diarrhea, and ARI among children under-5 were 36.92%, 4.84%, and 5.40% respectively. Multilevel modeling techniques showed that mother's age at birth, mother’s BMI, children's age, underweight of children, birth order of children was significantly associated with fever. Children's age and the child's chronic malnutrition or stunting were significantly associated with diarrhea. The study also determined children's age, sex, underweight, mother age and education as the influential risk factors of ARI. Conclusion: Comorbidity is common among the children of under-five in Bangladesh and affects girls and boys expressively. The child malnutrition is the major cause of child morbidity in Bangladesh. The study suggests that improving nutritional status is decreasing the comorbidity of children in Bangladesh. Keywords: Under-five, Prevalence, GLMM, Comorbidity, Bangladesh. Submission for Student Paper/Poster Competition? Yes

EC-096 District Level prevalence of C-section and Skilled Birth attendant in Bangladesh: An Application of Small Area Estimation Method Md. Ahadur Rahman1, Mossamet Kamrun Nesa1, Sabina Islam1 1Department of Statistics, Shahjalal University of Science and Technology, Sylhet, 3114, Bangladesh, [email protected], [email protected], [email protected]

National level indicators of maternal health often hide the real scenario across a country. In order to construct a maternal health indicators map, accurate estimates of the indicators are required at very small spatial scales, typically the administrative units of a country. Although comprehensive data on maternal health indicators are collected in national surveys, the small-scale estimates cannot be calculated using the standard estimation methods in national surveys since such methods are designed to produce national level estimates and assume large samples. Area level small area estimation methods (e.g., Fay-Herriot method)

147

are widely used to calculate such micro-level small-scale estimates. In this research, Fay-Herriot model has been used to calculate district wise estimates with mean squared error. The BDHS 2011 and Population Census 2011 are utilized for this study. The considered method has been applied to estimate prevalence of Skilled Birth Attendant (SBA) and C-section as indicators of maternal health at district levels with their accuracy measures. In terms of SBA, about 10 districts have lowest prevalence of SBA within the ranges of 0-16.3%. The scenario of C-section is also revealed, indicating that almost 12 districts have highest prevalence of C-section (26-35%) and the lowest was found in Bhola (5%). Keywords: Small Area Estimation, Fay-Herriot Model, SBA, C-section, Spatial Distribution Submission for Student Paper/Poster Competition? No

EC-097 Regional and country-level variation of hemoglobin level of women in South Asian countries: an evidence from 2011-2016 Demographic and Health Surveys Md Atiqul Islam1, Abdullah Al Islam2, Ruhul Amin3, Md Ahadur Rahman 1Department of Statistics, Shahjalal University of Science and Technology, Sylhet-3114, Bangladesh, Email: [email protected] 2Department of Statistics, Shahjalal University of Science and Technology, Sylhet-3114, Bangladesh, Email: [email protected] 3Department of Statistics, Shahjalal University of Science and Technology, Sylhet-3114, Bangladesh, Email: [email protected] 3Department of Statistics, Shahjalal University of Science and Technology, Sylhet-3114, Bangladesh, Email: [email protected]

Background: The hemoglobin (Hb) protein plays an essential role in health and development in the human body. Anemia results when Hb falls below necessary levels. Anemia is one of the crucial global public health headaches, especially in developing countries such as Bangladesh, India, Nepal, Maldives and Myanmar and several risk factors are responsible for it. Therefore, our aim is to identify the risk factors of low Hb concentration of South Asian countries women. Methods: The study extracted data from the recent five South Asian countries Demographic Health Surveys (DHS) conducted from 2011 to 2016. A total of 7,16,416 women were used in the study. Multilevel linear mixed effect modelling was applied to identify the risk factors of Hb level at different hierarchical levels simultaneously and the different level variations were observed. Results: The mean and SD of Hb level was 117.56±16.5 g/L in the overall situation. The highest mean of Hb level was found in the Nepal 123.0±15.4 g/L and the lowest in the Maldives 115.3±13.8 g/L. Multilevel model resulted that wealth index, education level, using contraceptive method, pregnant women, married women, breastfeeding, age, body mass index, total children ever born, access to piped water and presence of a toilet facility had an effect on the Hb level. Conclusion: The research highlighted the impact of the most important risk factors from the socioeconomic, demographic and environment on Hb level and indicating the tendency of anemic of women in overall the five South Asian countries. Keywords: Anemia, DHS, Hemoglobin, Multilevel, South-Asian

EC-098 Prediction of the Success of Android Apps before Launching on Google Play Store Golam Md. Muradul Bashir1, Md. Showrov Hossen2, Dip Karmoker3 and Md. Junaeed Kamal4 1Department of Computer and Communication Engineering, Patuakhali Science and Technology University, Dumki-Patuakhali, [email protected] 2Faculty of Computer Science and Engineering, Patuakhali Science and Technology University, Dumki-Patuakhali, [email protected] 3Faculty of Computer Science and Engineering, Patuakhali Science and Technology University, Dumki-Patuakhali, [email protected] 4Faculty of Computer Science and Engineering, Patuakhali Science and Technology University, Dumki-Patuakhali, [email protected]

Google play store, the mobile apps distribution platform gets flooded with several thousands of new apps every day. Developers work independently or in a team to make their apps successful. In the competitive world, it is important for a developer to know whether he is proceeding in the right way. Before launching apps on Google Play Store, the developers do not know whether their apps will be successful or not. The success of an app is generally determined by the number of installs and the user ratings. In this research,

148

analysing the large scale of google play store data, we have tried to predict the success of a new app before launching it into Google Play Store with enumeration of the apps installation and user rating. Keywords: android, mobile apps, rating, installation Submission for Student Paper/Poster Competition? Yes

EC-099 Application of Supervised Learning Methods for Gene Expression Microarray Data Classification Suman Khan, Bandhan Sarker, Priyanka Bosu and Md. Matiur Rahaman* Department of Statistics, Bangabandhu Sheikh Mujibur Rahman Science and Technology University, Gopalganj-8100, Bangladesh. Email: [email protected]; [email protected]; [email protected]; [email protected]

Abstract. In the clinical context, samples assayed by microarray are often classified by cell line or tumor type and it is of interest to discover a set of genes that can be used as class predictors. The leukemia dataset of Golub et al. and the colon dataset of Alon et al. (1999) are two cancer types were identified based on their origins. Leukemia data contain lymphoid (lymph or lymphatic tissue related) and myeloid (bone marrow related). ALL could be further divided into B-cell and T-cell ALLs. The number of genes in the study is 7128 with 72 samples where 47 ALL samples and 25 AML samples. Colon data set contain 62 samples with 2000 genes: 40 tumor tissues, coded 2 and 22 normal tissues, coded 1. We applied t-Statistics for determined differentially expressed genes. Then we get 746 genes and 186 genes are differentially expressed for 7128 leukemia dataset and 2000 colon dataset, respectively. Then we used 10-fold cross- validation and apply three different machine learning techniques (Support vector machine, Linear discriminant analysis and Random forest) and find its prediction accuracy. We iterate machine learning model 100 times and find its average prediction accuracy. We also used box plot for comparison. The conclusion is that those three methods patient prediction accuracy is more than 90% and 87% during the analysis of Leukemia and Colon dataset, respectively and Random Forest (RF) shows better prediction accuracy. Keywords: Machine learning, Microarray data, Support Vector Machine (SVM), Linear Discriminant Analysis (LDA), Random Forest (RF). Submission for Student Paper? Yes.

EC-100 Correlates of susceptibility to future tobacco use among adolescents in Bangladesh Shariful Hakim, Chander Hat Degree College and Md Jamal Uddin Shahjalal University of Science and Technology, Bangladesh

EC-101 Prediction of Malnutrition and Its Risk Factors among Under-Five Children in Bangladesh: A Comparison with Support Vector Machine and Logistic Regression S.M. Jubaidur Rahman1, N. A. M. Faisal Ahmed1, Md. Menhazul Abedin1 Benojir Ahammed1, Mohammad Ali1, Md.Maniruzzaman1, 2* 1Statistics Discipline, Khulna University, Khulna, Bangladesh 2Department of Statistics, University of Rajshahi, Rajshahi, Bangladesh *Correspondence: Md. Maniruzzaman ([email protected])

Background and Objective: Malnutrition is one of the most important health and welfare problems among under-five children in Bangladesh. Children are malnourished if the calories and proteins they take through their diet are not sufficient for their growth and maintenance. Adequate nutrition is essential in early childhood to ensure healthy growth, proper organ formation and function, a strong immune system, neurological and cognitive development. The main objectives of this study are to predict of malnutrition and its risk factors among under-five children suing using machine learning (ML)-based approach. Materials & Methods: This study utilized the data which extracted from the Bangladesh Demographic and Health Survey (BDHS), 2014 with 7079 sample size. Based on the previous literatures, we have selected 13 factors and 3 anthropometric indicators as response variable that are commonly used to describe the nutritional status of the children: weight-for-height (wasting), height-for-age (stunting) and weight-for-age

149

(underweight). Chi-square test is performed to show the association among dependent and independent variables. Then logistic regression (LR) model is used to identify the risk factors of malnutrition. Finally, two ML-based techniques as support vector machine (SVM) and logistic regression are used to predict the malnutrition status. Results: The overall prevalence’s of stunting, wasting and underweight are 13.8%, 15.4%, and 15.6%, respectively. It is observed that the prevalence of stunting, wasting and underweight is higher in rural area compared to urban. LR results demonstrates that 5 factors as division, age of children, father’s education, types of toilet and wealth index are the significant risks factors for stunting , 4 factors (division, age of children, sex of children and wealth index) for wasting and 5 factors (division, age of children, father’s education, types of toilet & wealth index) for underweight. It is noted that SVM with radial basis function (RBF) kernel gives the highest classification accuracy compared to LR. Conclusion: It can be concluded that an improvement in societal infrastructure, better education and nutrition are needed to address the child malnutrition issue. Keywords: Malnutrition, Bangladesh, logistic regression, support vector machine.

EC-102 A study on post-operative pain management and satisfaction after caesarean delivery in Rajshahi City. Farhana Hasan and Md. Golam Hossain Health Research Group, Department of Statistics, University of Rajshahi, Bangladesh

Background: After any obstetric surgery pain is a common discomposure. The aim of this study was to systematically asses the intensity of pain and ensuing patient satisfaction with pain management after caesarean delivery. Methods: A cross-sectional study was followed 195 women who underwent cesarean delivery in last five years. Patients were asked about duration of postoperative abdominal scar pain using the American Pain Society Patient Outcome Questionnaire (APS-POR-R) scale (0 to 10) of 24 hours after surgery. Satisfaction with pain treatment was ascertained at 24 hours after surgery using the same scale. Data were collected by using convenience sampling technique. Descriptive statistics (Mean ± SD) was utilized to find the level of severity of pain, pain management and patients’ satisfaction level. Statistical significance was accepted at p< 0.05. The severity and management of pain were classified into four classes; (i) No pain, (scale, 0), mild (scale, 1-3), moderate (scale, 4-7) and severe (scale, 8-10). Patients satisfaction also classified into four classes; (i) dissatisfied, (scale, 0), mild (scale, 1-3), moderate (scale, 4-7) and completely satisfied (scale, 8-10). Results: It was observed that the average pain in last 24 hours for least, worst and severe pain level were 6.94±2.82, 5.86±2.35 and 6.02±2.23 respectively. The average value of pain severity was moderate (6.27±2.47). The mean value of pain doing activities in bed was 4.66±2.95 while the mean pain of activities out of bed was 7.31±2.55. The adverse drug effects on caesarean post-operative women was measured by nausea, drowsiness, itching and dizziness, and the mean values of each items were 1.69±1.88, 2.15±2.74, 1.22±2.10 and 1.43±2.40 respectively. The average satisfaction level was 7.47±2.12and the average pain relief was found moderate (5.62±2.22) . Subjects were asked to rate their pain on 10-points numerical rate pain intensity was divided into three categories such as; 1-3 mild pain, 4-7 moderate pain and ≥ 8 severe pain. In the first 24 hours after surgery 20(10.3%) women reported mild pain, 122(62.6%) had moderate pain and 50(25.6%) women had severe pain. More than two thirds declared that the severe pain persisted fifty percent or less of 24 hours of surgery. Satisfaction score were high and 114(58.5%) women were extremely satisfied toward management of pain during hospitalization (score ≥ 7). Conclusion: In our study, the intensity of pain reported in the first 24 hours was quite high and women were satisfied with quality of pain management. Improving quality of pain management is multifaceted factors, which requires involvement of all aspects from care players, including nurses, patients and physicians who will follow a uniform guidelines and protocols. Keywords: Pain evaluation, postoperative, caesarean delivery, pain management, quality of daily life, decision tree.

150

EC-103 Correspondence Analysis to Patterning Heavy Smokers in Bangladesh Afia Ibnat1 and Papia Sultana2 1 Department of Statistics, University of Rajshahi, Bangladesh. E-mail: [email protected] 2 Department of Statistics, University of Rajshahi, Bangladesh. E-mail: [email protected]

Objective: This paper aimed to patterning heavy smokers in Bangladesh using correspondence analysis. Significance of the work: In Bangladesh, smoking is less likely to have priority in research and so far no works have been found related to heavy smoking. Therefore, this study will reveal pattern of heavy smokers in Bangladesh. Results: For the current study, data from Global Adult Tobacco Survey, 2010 of size 9619 had been used. Among the respondents 50.44% were from urban and 49.56% from rural area. It was also found among the respondents 46.40% were male and 53.60% were female. Multivariate correspondent analysis (MCA) showed that urban respondents mostly corresponded to non-smokers and rural respondents corresponded to light-smokers and non-smokers. Results also showed that male respondents more likely to be associated with light-smokers and heavy-smokers. It had been found that respondents of age group 35-44 years and 45-54 years were mostly associated to heavy smokers. Conclusion: In conclusion, it can be said that government and policy makers may modify or implement new policies to quit heavy smokers, especially for rural people, people with no formal education and with lowest household wealth index. Keywords: Heavy smoking, Correspondence analysis, correlates, GATS. Submission for Student Paper/Poster Competition? Yes

EC-104 Prevalence and predictors of heavy smoking in Bangladesh: Ordinal logistic regression approach Sharif Chowdhury Joy1 and Papia Sultana2 1Department of Statistics, University of Rajshahi, Bangladesh. E-mail: [email protected] 2Department of Statistics, University of Rajshahi, Bangladesh. E-mail:[email protected]

Objectives: Aim of the study is to estimate the prevalence of heavy smoking and to determine the significant predictors of heavy smoking using ordinal logistic regression model. Significance of the work: Bangladesh is one of the top ten countries in the world with high current smoking prevalence and the number of heavy smoker is increasing day by day. But, to best of our knowledge no research work has been found on heavy smoking in Bangladesh. Results: Data of size 9629 has been used from Global Adult Tobacco Survey, 2009. It has been observed that the prevalence of heavy smokers in Bangladesh is 5.22 (11.03% male and 0.19% female). From ordinal logistic regression analysis, it has been observed that respondents from rural area are 8% less likely to be heavy smoker compared with light and non-smoker than respondents from urban area (OR=0.92, 95%CI= 0.81, 1.06 and p-value=0.245). Male respondents are 33 times more likely to be heavy smoker compare with light and non-smoker than female respondents (OR=33.52, 95%CI= 22.85, 49.17 and p-value= <0.001). Respondents from age group 45-54 years are 3 times more likely to be heavy smoker than respondents of age 15-24 years (OR=3.68, 95%CI= 2.93, 4.62 and p-value= <0.001). College/University completed and higher respondents are 77% less likely to be heavy smoker than respondents with no formal schooling (OR=0.23, 95%CI= 0.17, 0.32 and p-value= <0.001). It has been also found that respondents from highest household wealth index are 47% less likely to be heavy smoker than respondents from lowest household wealth index. Conclusion: In conclusion, the results of this study reveal that heavy tobacco smoking is strongly associated with social disadvantage, for example, low socio-economic status, less education. Giving it as public health priority, WHO Framework Convention on Tobacco Control should be implemented for those disadvantaged group in priority basis. Keywords: Heavy smoking, predictors, ordinal logistic regression, GATS, Odds Ratio. Submission for Student Paper/Poster Competition? Yes

151

EC-105 BMI Based Nutritional Assessment of the Student of Varendra University Md. Nadim Sarwar1, Ahammad Hossin2 and Md. Ayub Ali3 1B.Sc. (Hon’s) student, Dept. of CSE, Varendra University, Email: [email protected] 2Assistant Professor, Dept. of CSE, Varendra University, Email: [email protected] 3Professor, Dept. of Statistics, University of Rajshahi, Email: [email protected]

All over the world health issue is the most important content of discussion. The main topic of this study is to calculate the percentage of student’s BMI with the classification by nutritional level of WHO 2019 among Varendra University from 358 students under 11 departments. This Study deals with the primary data of which was collected from face to face questionnaire survey through simple random sampling without replacement method from 5134 students under the 11 departments of university. Frequency distribution were conducted using IBM SPSS of version 23. This study reveals that there are 18.7% students have underweight, 68.2% students have normal weight, 10.3% students have over weight and 2.8% students are obese. Here we observed that 18.7% students are underweight. This can be happening by poverty or other illness. We should detect the problem to get remedy from this issue. Keywords: BMI, Nutritional Assessment, Student of Varendra University

EC-106 Water Uses in Rajshahi District: An Estimate Md.Shahanur Alam1, Md. Abdul Khalek2 and Md. Ayub Ali3 1 Research Student, Department of Statistics, University of Rajshahi, Email: [email protected] 2 Associate Professor, Department of statistics, University of Rajshahi, Email: [email protected] 3 Professor, Department of statistics, University of Rajshahi, Rajshahi 6205 E-mail: [email protected]

Water shortages, deterioration of water quality, ground water lowering, non-availability of surface water and environmental constraints, the government / water authorities need to consider and to invest in water conservation practices and waste water recycling. In this study, possible conservative use of water has been estimated through the population parameter estimating under multistage stratified random sampling. The study shows that the average water consumption for household activities was 85 LPCD in Rajshahi city. The houses having flushing toilet used more water than the houses having non flushing toilet. According to the study survey the water consumption for flushing toilet was 30 LPCD. Whereas the non-flushing toilet water requirement was 18 LPCD. Installing water-efficient plumbing fixtures, appliances contribute and habitual changes are the major tools for water conservation. The study shows that 28 LPCD water can be saved from toilet, bathing and faucets use by water conservation practices. Total estimated water was 143.6 MLD in Rajshahi city. Rajshahi WASA generates 95.00 MLD. In according to WASA total demand of water was 113.9 MLD in Rajshahi city. On average, total water use was 330.07 Liters Per Day per House hold in Rajshahi District and it was 408.64 liters in urban, 235.55 liters in rural and 321.00 liters in semi-urban. Government should take necessary action to save, reuse and not to waste our valuable water. Keywords: Water use, Water demand, Population parameter estimate

EC-107 A Comparative Study of Hierarchical Clustering Methods Based on Cancer Gene Expression Data Tapos Kumar Biswas1, Most. Farida Khatun2, N. A. M. Faisal Ahmed1 Md. Menhazul Abedin1, Benojir Ahammed1, Md. Maniruzzaman1, Mohammad Ali1* 1Statistics Discipline, Khulna University, Khulna, Bangladesh 2Pharmacy Discipline, Khulna University, Khulna, Bangladesh *Correspondence: Mohammad Ali (ali.ru. [email protected])

The clustering method is used for the discovery of cancer / subtypes have drawn a great attention in the scientific community. The medical community has preference for using hierarchical clustering method. There were no serious comparison studies in previous literatures using seven hierarchical clustering methods along with five proximity measures for microarray gene expression dataset. The main objective of this study is to identify the better combination of hierarchical clustering and proximity measures in presence and absence of outliers. In this study, we have used 15 microarray cancer gene expression datasets. We have

152 included randomly 5% and 10% artificial outliers in the datasets. We have used 35 combinations using seven hierarchical clustering methods as Single linkage (SL), Complete linkage (CL), Average linkage (AL), Ward’s, Median, Centroid, McQuitty along with five proximity measures (Euclidean distance, Manhattan distance, Pearson’s correlation coefficient, Spearman correlation coefficient, Cosine similarity). The performance of these of combination methods is evaluated using corrected rank (cR) index. Moreover, pair t- test is used to measure the difference in the mean value of cR compared with the large one based on p-value. Results indicate that McQuitty clustering technique with spearman correlation coefficient (MCQSP) slightly better perform to others. We suggest that in the community of bioinformatics McQuitty clustering technique with spearman correlation coefficient (MCQSP) should be used for microarray dataset clustering. Keywords: Cancer gene expression data, Hierarchical clustering, Proximity measures.

EC-108 Associated Factors of Exclusive Breastfeeding among Infant Aged 0-6 months in Bangladesh and its Effect on Diarrhea and Acute Respiratory Infection Sara khandker1, Mohammad Ali1, Md. Menhazul Abedin1 Benojir Ahammed1, Md. Maniruzzaman1, N. A. M. Faisal Ahmed 1* 1Statistics Discipline, Khulna University, Khulna, Bangladesh *Correspondence: N. A. M. Faisal Ahmed ([email protected])

Background and Objective: Exclusive Breastfeeding practices are effective ways for reducing childhood infectious diseases, morbidity and mortality. The objective of this study was to determine the prevalence and associated factors of exclusive breastfeeding (EBF) and its effect on diarrhea and acute respiratory infection (ARI) among infants aged 0-6 month(s) in Bangladesh. Methods: Data was extracted from Bangladesh Demographic and Health Survey (BDHS), 2014. BDHS, 2014 collected data from 17,863 Bangladeshi women in reproductive age from the entire country using two stages stratified cluster sampling. Here we included only 761 infants aged 0-6 month(s). The association between the independent variables and outcome variables were evaluated using Chi-square test. Logistic regression was used to determine the significant risk factors related to EBF and its effect on diarrhea and ARI. Results: The overall prevalence of EBF among infants aged (0-6) month(s) were 387 (50.9%of total sample) at the time of interview, and 374 infants (49.1%) were never exclusively breastfed. Among the covariates only four factors (child age of month, child twin, region, toilet facilities) were significantly associated (p<0.05) with exclusively breastfeeding. The data suggested that for 0-6-month(s) aged infants the prevalence of diarrhea and acute respiratory infection was 14.3% and 31.2% respectively. The prevalence of both illnesses was significantly associated with lack of exclusive breastfeeding. A binary logistic regression indicated that relatively less EBF practice infants were more suffering from diarrhea [AOR=0.446, 95% CI: 0.214-0.929, p<.05] and acute respiratory infection [AOR=0.532, 95% CI: 0.392-0.721, p<0.05] with EBF. Conclusion: A protective effect of exclusive breastfeeding against infectious diseases-related morbidity in infancy and showed that frequently-collected cross-sectional datasets could be used for estimating effects. The low prevalence of exclusive breastfeeding in Bangladesh needs to be improved to decrease child morbidity. Keywords: Acute respiratory infections, Exclusively breastfeeding, Diarrhea, Bangladesh.

EC-109 Identification the similar genes of NOG1 and the evolutionary relationship using protein sequence data Tahira Mahbub1, Most. Farida Khatun2, Md. Menhazul Abedin1 Benojir Ahammed1, Md. Maniruzzaman1, N. A. M. Faisal Ahmed1, Mohammad Ali1* 1Statistics Discipline, Khulna University, Khulna, Bangladesh 2Pharmacy Discipline, Khulna University, Khulna, Bangladesh *Correspondence: Mohammad Ali ([email protected])

Background and Objective: The concept of a gene continues to be refined as new phenomena are discovered. It is the functional unit of inheritance controlling the transmission and expression of one or more traits. NOG1 gene increase rice production. NOG1 gene is a nuclear binding protein present in eukaryotic

153

cell. The objective of this study was to identify the similar gene of NOG1 and their evolutionary relationship using protein sequence data. Materials and Methods: We took 73 query sequences and one test sequence. Blast online software was performed to identify the similar genes. ClustalW and MUSCLE were used for identifying best alignment. Then we performed disparity index test to observe the heterogeneity. Finally we conducted phylohenetic analysis to observe the evolutionary relationships. Results: From the BLAST online software we got 23 similar genes with their E values. Then we performed MSA methods. One is ClustalW and other is MUSCLE. From the result we found MUSCLE was a good MSA method. We performed disparity index test and the result showed us homogeneity among the genes. At last phylogenetic analysis was done and from the tree we could observe that the most similar genes were situated in nearest ancestor. Oryza sativa indica group was our query sequence and after the phylogenetic analysis we got Oryza sativa japonica group share the nearest ancestor. So Oryza sativa japonica group could be a good replacement of Oryza sativa indica group. Conclusion: We found Oryza sativa indica which contain NOG1 gene are similar with 3-hydroxyisobutyryl- CoA hydrolase like protein 5 oryza sativa japonica group. So we could say that rather than NOG1, if a paddy plant contains 3-hydroxyisobutyryl-CoA hydrolase like protein 5 gene it may help to increase rice production. Or if plants contain both these genes it also may help to produce high amount of rice. Keywords: Oryza sativa, NOG1 gene, Disparity index, Evolutionary relationship.

EC-110 Classification Based on Multinomial Logistic Regression Model: A Case Study on Women Empowerment Issue in Bangladesh Antora Das1, Ananya Paul Powshi2, and Md. Rezaul Karim3 1Department of Statistics, University of Rajshahi, Email: [email protected] 2Market Systems Specialist, Feed the Future Bangladesh Rice and Diversified Crops Activity, ACDI/VOCA, Email: [email protected] 3Department of Statistics, University of Rajshahi, Email: [email protected]

Classification is a broad class of models/methods related to categorization. It is useful for the prediction of a response (or outcome) for given values and/or levels of one or more independent (or predictor) variables. Multinomial logistic regression is one type of model that can be applied for classification purpose. This paper investigates the women empowerment issue which has a significant effects on the overall development of Bangladesh. It considers the Bangladesh Demographic and Health Survey (BDHS) 2014 data set for assessing Bangladeshi women’s empowerment status at household level. Here, the decision making on using contraceptive method is considered as an indicator of empowerment of women. This response variable is modeled depending on a set of predictor variables by using the multinomial logistic regression as a classification algorithm. The performance of this algorithm for women empowerment determination is measured in terms of sensitivity, determining prediction, and misclassification error rate. The research may be a boon to create expert decision making systems and a solution for women rights and decision making policy to construct an optimal prediction model for women empowerment. Keywords: Multinomial Logistic Regression, Women empowerment, Classification, Accuracy, Prediction. Submission for Student Paper/Poster Competition? Yes

EC-111 Divide and Recombine Approach for Analysis of Failure Data Using Parametric Regression Model Md. Razanmiah, Md. Kamrul Islam and Md. Rezaul Karim Department of Statistics, University of Rajshahi, Bangladesh E-mail: [email protected]; [email protected]; [email protected]

The failure data of some products depend on factors or covariates such as the operating environment, usage conditions, etc. Under this situation, the parametric regression model is applied for modeling the failure data of the product as a function of covariates. Divide and recombine (D&R) is a new statistical approach to the analysis of big data. In D&R approach, the data are divided into manageable subsets, an analytic method is applied independently to each subset, and the outputs are recombined. This paper applies the D&R approach for analysis of an automobile component failure data using the Weibull regression model. Extensive

154

simulation studies are presented to evaluate the performance of the proposed methodology with comparison to the traditional statistical estimation method. Keywords: Big data; Divide and recombine (D&R) approach; Failure data; Weibull regression model. Submission for Student Paper/Poster Competition? Yes

EC-112 Challenges and Methodologies to Handle Big Data Hamida Muksita Mustari, Jenarul Islam, Md. Limon Sharkar, Md. Rezaul Karim Department of Statistics, University of Rajshahi, Bangladesh

In the information era, enormous amounts of data have become available on hand to decision makers. The world’s data collection is reaching a tipping point for major technological changes that can bring new ways in decision making, managing our health, cities, finance and education. While the data complexities are increasing including data’s volume, variety, velocity and veracity, the real impact hinges on our ability to uncover the `value’ in the data through Big Data Analytics technologies. Big Data Analytics poses a grand challenge on the design of highly scalable algorithms and systems to integrate the data and uncover large hidden values from datasets that are diverse, complex, and of a massive scale. A number of major challenges appears in the analysis of big data. This poster is an effort to present current challenges of dig data analysis and existing methodologies to handle big data problems. Submission for Student Paper/Poster Competition? Yes

EC-113 Divide and Recombine (D&R) Method on Big Data Jenarul Islam, Md. Limon Sharkar, Hamida Muksita Mustari, Md. Rezaul Karim Department of Statistics, University of Rajshahi, Bangladesh

Big data is an evolving term that describes a large volume of structured, semi-structured and unstructured data that has the potential to be mined for information and used in machine learning projects and other advanced analytics applications. Big data is often characterized by the 3Vs: the extreme volume of data, the wide variety of data types and the velocity at which the data must be processed. More recently, several other Vs have been added to describe big data, including veracity, value and variability. Although big data doesn't equate to any specific volume of data, the term is often used to describe terabytes, petabytes and even exabytes of data captured over time. Nowadays Computational performance is challenging because datasets are huge in size, computational complexity of analytic methods is high, and computer hardware power is limited. When the computations have high complexity then small datasets can be challenging too. Divide and recombine (D&R) method is a statistical framework deals with these challenges under the analysis of big data. In D&R method, the data are divided into subsets. Then any numeric and visualization method, which collectively is analytic method, is applied to each subset independently. For each analytic method, the outputs of the application of the method for the subsets are recombined for full dataset. So each analytic method has associated with a division method and a recombination method. This poster will show the pairs of divide-recombine procedures that provide good results for analysis of big data using D&R method. Submission for Student Paper/Poster Competition? Yes

EC-114 An Overview of Class Imbalance Problem Md. Limon Sharkar, Jenarul Islam, Hamida Muksita Mustari, Md. Rezaul Karim Department of Statistics, University of Rajshahi, Bangladesh

Class imbalance problem is one of the major challenges of data science fields. Imbalance data sets degrade the performance of data mining and machine learning techniques as the overall accuracy and decision making be biased to the majority class, which lead to misclassifying the minority class samples or furthermore treated them as noise. This poster is an effort to present a general overview of class imbalance problem solutions and the most significant investigations recently introduced by researchers. Submission for Student Paper/Poster Competition? Yes

155

EC-115 An Overview of the Statistical Analysis of High-Dimensional Data Md. Limon Sharkar, Jenarul Islam, Hamida Muksita Mustari, Md. Rezaul Karim Department of Statistics, University of Rajshahi, Bangladesh

Technological innovations have revolutionized the process of scientific research and knowledge discovery. The availability of massive data with low cost, increasing data storage capacity and advancement of computational power have reshaped statistical thinking, data analysis and theoretical studies. Thus, modern applications of statistical theory and methods can involve extremely large datasets, often with huge numbers of measurements/features (p) on each of a comparatively small number of observations/experimental units (N) i.e. . New methodology and accompanying theory have emerged in response. This overview introduces statistical challenges with high dimensionality and their recent solutions. Submission푝 ≥ for푁 Student Paper/Poster Competition? Yes

156

INVITED & CONTRIBUTED Through E-mail

7th International Conference on Data Science and SDGs: Challenges, Opportunities and Realities

EM-001 Erlang Loss Formulas: An Elementary Derivation Jyotirmoy Sarkar Department of Mathematical Sciences Indiana University - Purdue University Indianapolis, IN 46202, USA Email: [email protected]

Discovered in 1917, the celebrated Erlang loss formulas express the probability that out of c identical channels/servers exactly j servers are busy serving customers. Today, let us ask: “What is the simplest proof of these formulas?” As an alternative to more advanced methods, we derive the Erlang loss formulas using (a) an intuitive limit theorem of an alternating renewal process, and (b) recursive relations that are solved using mathematical induction. Thus, we make the Erlang loss formulas comprehensible even to beginning college mathematics students. Furthermore, we illustrate how to make decisions in some practical problems using these formulas and other quantities derived from them. Keywords: Cycle time, ergodicity, queuing system, renewal time, semi-Markov process 2010 MSC Code: Primary 60J28; 60K20, Secondary 62-01

EM-002 Prevalence of over nutrition, anaemia and other non-communicable diseases among adult Indians Premananda Bharati, Manoranjan Pal and Susmita Bharati Indian Statistical Institute, 203 B.T. Road, Kolkata 700108, India

Overweight or obesity is a big problem among Indians like other Asian countries, though its intensity varies due to interstate disparity in socio-economic and socio-cultural differences. Besides over nutrition, there is a great disparity of prevalence of anaemia and various other non-communicable diseases like Diabetes, Asthma, Thyroid disorder, Heart diseases, Cancers, etc. Thus, it is worth to study state wise variation of over nutrition and anemia and the noncommunicable diseases among Indians, by gender and also by socio- economic variables. For this study, we have used data of adult Indians of age 15-49 years taken from NFHS- 4. The socio-demographic variables like age-groups, type of residence, marital status, educational status, religion, wealth index etc. and Body Mass Index (BMI) have been taken to see the effect of these variables on the morbidity patterns. The study reveals that in India, percentage of over nutrition among women is 20.7 and among men it is 18.9. The gender difference is only 1.8%. Percentage of anemia among adult women is 53.0 and among men is 22.7, the gender difference being 30.3%, i.e. females are much more affected than males. Regarding other non-communicable diseases, percentage of Diabetes among women and men are found to be same but percentages of Asthma and Thyroid are found to be more among women than men though there is regional variation. Percentages of women and men affected by heart disease and cancer are more or less same, though women are a bit more affected than men. Over-nutrition concentrated zones in India are North, West and South whereas high anaemic affected zone is the east zone. The highest concentrated zone for many of the non-communicable diseases is found to be the south zone. If we compare our results of the present study with those of NFHS-3 data, it is seen that over nutrition has increased by 8.1% for women and 9.6% for men, whereas anaemia has decreased by 2.3% for women and 1.5% for men over the last ten years. The possible socio-economic factors responsible for over nutrition and anaemia have been identified and it is discussed in detail in the analysis section of the paper.

EM-009 Recent breakthroughs and anticipation in new-born phenylketonuria screening and management program in Bangladesh Md. Abdul Alim Al-Bari* Department of Pharmacy, University of Rajshahi, Rajshahi-6205, Bangladesh * Corresponding Email: [email protected]

New-born screening aims at the earliest possible recognition of disorders to prevent the most serious consequences by timely intervention. This screening needs a confirmatory diagnosis and requires further investigations. Guidelines from some developed countries recommend new-born screening as soon as possible after birth before hospital/clinic discharge because of the high prevalence of certain endocrinopathies e.g., congenital hypothyroidism (CH) and metabolic errors e.g., phenylketonuria (PKU) which, if recognized later, contribute to significantly increase in disability as well as morbidity. Although the

159

exact list differs among, and sometimes within countries, testing for PKU and hypothyroidism is universal in the developed world. Unfortunately, neonates are not screened in Bangladesh because the health policies have typically targeted mortality and infectious morbidities but not congenital disabilities. These former policies have been successful in lowering infant mortality rates, but the net effect of these gains has been somewhat offset by an increase in disability. The study aims to investigate the current status PKU under new-born screening (NBS) program in the northern region of Bangladesh, as an example of a developing country, focusing also on future plans to manage with life-long treatment. The primary challenges are identified and implemented for national PKU-NBS policy as a role model of Bangladesh for developing countries. Key words: Phenylketonuria, new-born screening, Guthrie test, Bangladesh

EM-010 Characterisation along with stability for the wrapped exponential distribution using adjusted rate of failure Rishideep Roy Assistant Professor, Indian Institute of Management Bangalore Email: [email protected]

The exponential distribution, which provides with an important asymmetric model for describing life distributions on a linear domain, can be wrapped on a unit circle for modeling circular data. However, the wrapped distribution, so generated, fails to retain the constancy of the failure rate property. The present endeavour is to suitably adjust the failure rate function for wrapped and truncated distribution, and characterise the wrapped exponential distribution based on constancy of the adjusted rate of failure. Keeping in mind the empirical requirement, we have examined the stability of this characterisation property based on small departure from the state of constancy of the adjusted rate of failure.

EM-011 Two sample testing of proportion: Non-Null case Shubhabrata Das Decision Science Area, Indian Institute of Management Bangalore, India 560076

Consider the traditional two sample testing of hypothesis of proportion with the null hypothesis being : = , based on (large) independent samples (of sizes n1 and n2) drawn from the two populations.퐻0 휋1 − 휋2 In 𝛿𝛿the0 null case (i.e. for = 0), the (Z) test statistic uses standard error (1 )( + ) ∗ ∗ 1 1 0 computed from the pooled estimate𝛿𝛿 of proportion = , where p1, p2 �are푝 the− 푝two 푛sample1 푛2 푛1푝1+푛2푝2 proportions. This is appealing as the pooled estimate p* ∗is the MLE of π = π . However, in the stated non- 푝 푛1+푛2 1 2 null case, = 0 similar approach is typically not adopted, possibly because of lack of close-form solution of the MLE estimate. However, this is a weak justification with present day ease of computational facility. 0 While the 𝛿𝛿adopted procedures may be asymptotically equivalent, its performance may be worse for moderate sample sizes. In this work, we discuss computation of restricted MLE of the population proportions under null hypothesis with = 0, and compare the different standard errors for the various sample sizes to draw an insight for potential loss of efficiency in not using the MLE. Simple implementation for individual problem is shown 0 via MS𝛿𝛿 -Excel Solver, while general performance in complete set of possibilities for sample sizes up to 1000 are studied in details using R to characterize the comparative performance. Results exhibit a strong case for using restricted MLE under non-null case in some situation. Different applications of such problems will be discussed. The treatise may also be useful for computation of power in the null case.

160

EM-013 Implementation Status of Sustainable Development Goals (SDGs) in Bangladesh: A Statistical Approach Dr. Muhammad Amir Hossain1, Shehan Uddin Khan2 and Md. Sharful Islam3 1Deputy General Manager, Bangladesh Bank, Dhaka, Email: [email protected] 2Joint Director, Bangladesh Bank, Dhaka, Email: [email protected] 3Student, Department of Physics, Rajshahi University, Email: [email protected]

Head of member countries of United Nations (UN) agreed to move forward from the Millennium Development Goals (MDGs) to the Sustainable Development Goals (SDGs), a "One World" approach to the global development agenda with 17 Goals, 169 Targets and 232 indicators to be achieved by 2030. key objectives of these goals are to: end poverty and hunger, ensure healthy lives and promote well being, ensure inclusive and equitable quality education, promote peaceful and inclusive societies, provide access to justice for all and build effective, accountable and inclusive institutions at all levels. Bangladesh started work to achieve all goals within timeframe of UN Proportion of population below international poverty line, prevalence of undernourishment & stunting, maternal mortality rate, literacy rate, proportion of women in managerial positions, access to clean & sanitation services, access to electricity, growth of GDP per capita are in track to achieve goal by 2030. Government has adopted many policies and programs to address multidimensional nature of poverty, to ensure food security and to address specific hunger and nutrition issues in the country. As Bangladesh achieved successfully many indicators of MDGs, in the same way hopefully Bangladesh will success significantly to achieve SDGs. Keywords: MDGs, SDGs, Statistics, Sustainable Development, Targets

EM-014 Impact of Socio-Economic Condition on the Nutritional Status of School Girls An Anthropometric Approach Barnali Chakraborty Ph.D Student, West Bengal State University, Barasat, West Bengal [email protected]

Introduction: Nutrition and health were the most important contributory factors for human resource development in the country (Amirthaveni and Barikor, 2001).The large section of Indian population were suffered from varying degrees of protein energy deficiency (Gopalan, 2002). Malnutrition is still considered one of the major public health problems in many countries, affecting more than 30% of children under 5 years of age and the most vulnerable group was preschool children. Even now under nutrition in the young children is one of the public health concerns especially in the developing countries (world Bank ,2000).The main objective of the present study was to find out the effects of different socioeconomic indicators on the nutritional status of the school going children(girls) belonging in two different economic group. Methodology: This Cross-Sectional study was conducted in two school (Govt. & Private school of Ashokenagar and Guma village), North 24Parganas of West Bengal and children of the age 6 to 9 were measured from both the schools. The children are selected randomly for necessary anthropometric measurements and socio economic informations (family income, parents educational level) are collected from school’s register and the age are calculated on the basis of this date of birth. Result: The study concludes with a general statement that economy has a pivotal role in health and nutritional status of children under the age of 6 to 9ss years .The extent of these imbalance at higher age require to be investigated. Since growth velocity starts almost individually at the age of nine years, it is more critical to assess the impact of economy on the physical development of children. Keywords: Nutrition, primary school child.

161

EM-015 Measurement of Food Insecurity and Poverty by Different Methods - Does the Method Matter? Dr. Md. Zakir Hossain Professor, Department of Statistics, Shahjalal University of Science & Technology, Sylhet, Bangladesh E-mail: [email protected]

A variety of methods was found to use to measure the level of food insecurity and poverty. Among the methods of measuring food insecurity and poverty, coping strategy index (CSI) score, perception analysis, food consumption score (FCS), costs of basic needs (CBN), direct calorie intake (DCI), and food energy intake (FEI) were widely used. It is essential to verify whether the incidence of poverty and the level of food insecurity vary due to selection of measurement methods. This study aims to explore whether method selection does matter in evaluating the level of food insecurity and poverty. The data for this study has been collected through a research project “Reducing the Errors in the Measurement of Food Security in Bangladesh through Development and Implementation of an Innovative Method”, sponsored by the Grants for Advanced Research in Education (GARE) Program of Ministry of Education, Government of Bangladesh. The study has adopted cluster-sampling methods in which the primary sample units of Bangladesh Bureau of Statistics (BBS) have been considered as clusters. Based on the recognized sample- size determination formula, data have been collected from 600 sample households covering 30 rural clusters. This study has performed analysis on the complete data of 538 households. The CSI score method reveals that 21.2% of the surveyed households are found to suffer from normal food insecurity, 4.5% from moderate food insecurity and 5.4% from severe food insecurity. The perception analysis explores that about 18% of the households appeared to have normal food insecurity, 9.7% moderate food insecurity, and 3.5% severe food insecurity. The Food Consumption Score (FCS) method finds out that 5.6% of the surveyed households suffered from severe food insecurity (score 28 or less), about 8% from moderate food insecure, and 16.2% from normal food insecurity. The study findings indicates significant variation among the estimates of food insecurity by different methods. Based on CBN method, it is found that 12.3% households lie below the lower poverty line and about 22.1% below the upper poverty line. Based on the food intake data collected through measurement, it is found that 15.24% households lie below hardcore poverty and 31.78% households lie below absolute poverty. The modified approach (households took less amount of food than required considering age and sex of the members) of DCI method provides that 25.7% households lie in poverty. The incidence of poverty has been estimated by the FEI method considering average, weighted average and weighted regression model. The incidence of poverty is estimated at 16.4% for simple average method using the cut-off point of per capita monthly expenditure. However, it rose to 15.1% and 18.0% for weighted average methods where weight has been assigned by population density and asset score respectively. The incidence of poverty is found lower when weighted regression model has been applied. The overall findings indicate that the incidence of poverty varied from 13-18% by FEI method. Since this study explores significant variations among the estimates by different methods, method selection does matter in estimating the incidence of poverty and the level of food insecurity.

EM-016 The efficient way of crossing the row limit, while exporting big SAS® data to Excel Md. Mizan Alam1, Sirajul Islam2 and Md. Jahangir Alam3 1 SAS Programmer, AB Science, 3, Avenue George V 75008 Paris, France) 2 Senior Statistical Programmer, Shafi Consultancy Bangladesh, Bangladesh 3 PhD Fellow, Bioinformatics Lab, Department of Statistics, University of Rajshahi, Rajshahi – 6205, Bangladesh *Presenting Author ([email protected])

Microsoft Excel spreadsheets are often the format of choice for the users, both when supplying data to the processes and as a preferred means for receiving processing results and data. SAS® offers a number of means to import excel data quickly, efficiently and the methods are equally flexible to move data and results from SAS to excel. You know the row limit of xls and xlsx workbook is 65,536 and 1,048,576 respectively. While exporting the big data into excel, SAS generate an error message in the log file and the sheet cannot contain the rows above the limit, when it crosses the row limit. This paper will outline the technique to overcome the issue by splitting the big SAS data into multiple excel sheets automatically. Also, the data will be maintained cluster wise in every sheet i.e. if there have different

162

groups in the data volume, it will not split from one sheet to other. Thus, this will be a very efficient tip of moving data and results from SAS to excel. Keywords: Big Data Limitations, XLS-XLSX, EXPORT, ODS, SAS 9.4.

EM-017 A study on Blood Pressure among Young Adult of three Particularly Vulnerable Tribal Groups (PVTGs) of West Bengal, India Ramendra Nath Kundu1, Subir Biswas2 and Premananda Bharati 1Senior Research Fellow (UGC-NET) Department of Anthropology, West Bengal State University, West Bengal, India, Email: [email protected] 2Professor, Department of Anthropology, West Bengal State University, West Bengal, India 3Former Professor& Head, Biological Anthropology Unit, Indian Statistical Institute, Kolkata 700 108, India

Background: India is the second largest concentration of tribal population in the world and among them, 75 tribal communities are classified as Particularly Vulnerable Tribal Groups (PVTGs) on the basis of their primitive technology, population growth, poor literacy and economic condition. But at the same time, blood pressure is increasing among them. Objective: The study intends to find out the blood pressure condition among the young adult PVTGs of West Bengal and to find out the responsible potential factor of hypertension among the studied populations. Materials and Methods: It is a cross sectional study and is concentrated among Birhor, Lodha and Toto community of West Bengal. Sample size is 809 adult (18-40 years) male and female. Data was collected through certain anthropometric measurements, skin fold thickness and blood pressure through universally accepted standard techniques. Results: In these three PVTGs, percentage of normal blood pressure category is higher. Prevalence of high blood pressure found among the Toto community. In this study, Central obesity or Waist Height Ratio (WHtR) is the only reason for the increasing trend of blood pressure. Stepwise multiple regression coefficients shows that the beta values are highly significant (p<0.05 and p<0.01). Key words: Particularly Vulnerable Tribal Groups, Systolic Blood Pressure, Diastolic Blood Pressure, Body Composition Indicators, Central obesity.

EM-018 Assessment of nutritional status of a Particularly Vulnerable Tribal Group in India: A study on Adult Birhor community of West Bengal, India Nagma Parvin1, RamendraNath Kundu2 1Junior Research Fellow (UGC-NET) Department of Anthropology, West Bengal State University, West Bengal, India, Email: [email protected] 2Seniour Research Fellow (UGC-NET) Department of Anthropology, West Bengal State University, West Bengal, India, Email: [email protected]

Objective: The present study intends to find out the nutritional status among the adult Birhor community of West Bengal, India. Materials and Methods: A cross sectional study has been carried out among 164 adult Birhor tribes (aged 18 – 60 years) from Purulia district, West Bengal. Area was selected through deliberate sampling method whereas, total enumeration of participants has been taken, and data were collected after oral consent. Some anthropometric measurements were collected, as well as Body mass index (BMI), Percentage of body fat (PBF), Fat mass (FM), Fat free mass (FFM), Waist hip ratio (WHR) and Waist height ratio (WHtR) were calculated for the determination of nutritional status. Results: Nutritional status is different between male and female participants of the studied population. Among male participants 33.93 percent are underweight and among females 49.32 percent are reported as underweight. Result indicates that Fat mass and Percentage of body fat indicates female Birhors have more body fat compare to the male Birhors, on the other hand, in case of fat free mass male participants have less fat and more muscle compare to the female participants. Key words: West Bengal, Birhor , Nutritional status, Body Mass Index, Fat Mass

163

EM-020 Unequal Effect of Parents’ Education on Child Health and Health Care Manoranjan Pal , Utpal Kumar De, Premananda Bharati and Susmita Bharati Economic Research Unit, Indian Statistical Institute, 203 BT Road, Kolkata 700108, India Email: [email protected]

Parents have tremendous influence on growth and development of children. At the initial period of life, it is believed that mother is more important because there are some physical needs of children e.g., breastfeeding, which can only be done by mother. Both the parents should have emotional touch with the child. Apart from breastfeeding, it is generally believed that mother plays more vital role than father on the life of her child. Are there scientific evidences to support the belief that mother has more influence on child? We have not yet come across any such investigations other than taking opinion surveys. While we do not go against these opinions, we emphasize on the need of judging it scientifically. The paper carries out a systematic way of finding the effects of education of parents and their relative importance on the life of a child so far as Birth Weight, BCG, DPT, Polio and Measles Vaccination and Status of Anemia are concerned using unit level data of the latest round of National Family Health Survey (NFHS IV, 2015-16) of India. State wise secondary data are also used to see the effect of literacy of parents on the Sex Ratio at Birth of Children. Apart from preparing two-way tables, statistical tools like correlation, partial correlation and regression techniques are used for this purpose. Results of our analysis suggest that there is really unequal effect of mother and father and it is the mother, who counts more. Key Words: Educational Levels of Parents, NFHS, Birth Weight, Sex Ratio at Birth, Vaccination, Status of Anemia

EM-023 Nutritional Status of Rural Women: A Human Rights Concern Madhuparna Srivastava1, Susmita Bharati3, Manoranjan Pal1 & Premananda Bharati2 1 Economic Research Unit, 2 Biological Anthropology Unit 3 Sociological Research Unit, Indian Statistical Institute, 203, B.T. Road, Kolkata-700108

India is predominantly a rural country where, 72.4% workforce resides in the rural areas. Traditionally, agriculture is the prime sector of rural economy and rural employment. Women comprise a sizeable section of the rural workforce. Yet their potential contribution remains constrained by inequality and other aspects. One of the Sustainable Development Goals (SDGs) deals with the issue of gender inequality. It also recognizes the critical role women play in the achievement of all the goals. In this context, the paper intends to focus on the association of socio-demographic and economic variables with the nutritional status of women above 18 years of age. The data for this study has been obtained from a data set collected through primary survey as a part of a larger research work of the author. The study is based in the rural areas of the Bardhaman district of West Bengal. For assessing the nutritional status of women, Body Mass Index has been considered. We have taken data on socio-demographic and economic variables to see how these variables affect BMI. Though we are constrained by the size of the data, still we have found significant association between occupation and nutritional status. The paper also discusses the possibilities of integrating human rights-based approach, in combating the issues at hand. Keywords: Rural Women, Nutritional Status, Sustainable Development Goals, Human Rights

EM-025 Child Health in Five Tribal Societies of West Bengal, India TilakBagchi Former Assistant Keeper, Anthropological Survey of India, Western Regional Centre, Udaipur E-mail: [email protected]

Population stabilisation is crucial to sustainable development.If the people specially the children, who are the most vulnerable segment of population, are not fed adequately and live in unhygienic living conditionsuffer from illhealth owing to synergistic interaction between infection and water as well as vector borne and associateddiseases. As a result of the same their entire panorama of development would likely to

164

becompletely arrested. Poor health status and under nutrition thus is really an impediment of development. Against the above backdrop in the present treatise an attempt has been made to study health and nutritional panorama and associated cultural background of five tribal communities, viz, the Lodha,Mahali, Kora,Munda and Santal of PaschimMedinipore,West Bengal, India. The present study shows traditionally pre – agricultural tribal communities namely the Lodha, Mahaliand Kora have relatively poor health and nutritional status than that of the traditionally agricultural tribal communities namely the Munda and Santal. Cultural practices associated with the health and nutritional scenario of the tribal communities stated above have also been dealt in the present paper in order to highlight bio-cultural determinants responsible for the health status of the children of five tribal communities stated above.

EM-026 Temporal trend of malnutrition among adult women in North-east India: A study through two recent national level data Susmita Bharati, Manoranjan Pal and Premananda Bharati Indian Statistical Institute, 203 B.T. Road, Kolkata 700108, India

Improved nutrition of women is the base for development in health, education, employment, empowerment and poverty reduction as a whole. The main objective of the paper is, thus, to study the nutrition profile of reproductive aged women of North-east India and its changes through two recent successive national level data and to see the association of it with different socio-economic variables and certain non-communicable diseases. We have also seen the changes of the relevant parameters between two recent periods namely 2005-06 and 2015-16. The sample sizes are 18,752 (NFHS-3) and 91,694 (NFHS-4) of ever-married, non- pregnant reproductive aged women of eight states of North-East India. Dependent variable is Body Mass Index (BMI) and independent variables are place of residence, level of education of women, religion, ethnic status, wealth index of the family and certain non-communicable diseases. The finding shows that the highest prevalence of underweight and overweight states are Assam and Sikkim respectively as seen through the most recent NFHS-4data. Regarding temporal trend, in all the states of north-east India excepting Arunachal Pradesh, prevalence of underweight has been decreasing and that of overweight has been increasing from its previous study period. Poorer and middle-class women have become more obese for ten years gap. It is also to be noted that magnitude of obesity has decreased. Magnitude of underweight or obesity is changing more among middle class and poorer class. Does that mean that the levels of the bottom classes have increased over the time? In other words, is India approaching to developed countries? Regarding prevalence of non-communicable diseases, percentage has increased remarkably in NFHS-4 compared to NFHS-3. Thus, it can be said that, in north-east India, not all the states are equally affected, and many socioeconomic and cultural factors are associated with the level of BMI. But, economic condition is one of the important factors, which can improve the BMI level uniformly in all the cases.

EM-027 Determinants of Households’ Adaptation Practices against Climate Change Impact on Off- Farm Activities in Nepal Ananta Raj Dhungana1, Vikash Kumar KC2, and Purna Bahadur Khand3 1School of Development and Social Engineering, Pokhara University 2Department of Statistics, Tribhuvan University, PN Campus 3School of Business, Pokhara University

Climate change is a global challenge facing humans and their socio-economic activities, health, livelihood, and food security. Although adaptation practices against impact of climate change have been explored into practice in massive scale, the impact of climate change in agriculture production is most challenging. This study aims to analyze the determinants of households’ adaptation practices against climate change impact on off-farm activities in Nepal. It utilizes the data generated from nationally representative samples of National Climate Change Impact Survey, 2016 conducted by Central Bureau of Statistics. Total sample size of this survey was 5060 households. But for this paper, total of 4114 samples were considered. Binary logistic regression analysis was carried out to analyze households’ adaptation practices against climate change

165

impact on off-farm activities in Nepal. Most of the respondents are males of age group 40-54 years, from non-Brahmin/Chhetri caste/ethnicity, illiterate, with lowest income Quintile, from tropical climate zoneand without getting any services from agricultural service center. Age does not matter for determining the households’ adaptation practices towards all selected off-farm activities. Female is less likely to have adaptation practices towards off-farm activities (started more off-farm activities; shifted to non- agricultural employment; and temporary out-migration) compared with male while it does not matter for changed on food consumption habit. Non-Brahmin/Chhetri caste/ethnicity with reference to Brahmin/Chhetri is the determining factor for the adaptation practices (shifted to non-agricultural employment and temporary out-migration). Status of receiving any services from agricultural service center, years of experience in agricultural sector and sub-tropical climate zone with reference to tropical zone are the common determining factors for households’ adaptation practices towards off-farm activities against climate change impact in Nepal. Keywords: Adaptation, climate change, determinants, off-farm based activities, binary logistic regression

EM-029 Exclusive breastfeeding and incidence of acute lower respiratory tract infections in the first two years of life in children of a Bangladeshi slum Md. Ashraful Alam1*, Mustafa Mahfuz1, Tahmeed Ahmed12 1icddr,b, 68, Shaheed Tajuddin Ahmed Sarani, Mohakhali, Dhaka 1212, Bangladesh 2JPG School of Public Health, BRAC University, Mohakhali, Dhaka 1212, Bangladesh *Md. Ashraful Alam, Statistician, icddr,b. Email: [email protected]

Background and Aims: Acute lower respiratory tract infection (ALRI) – a major cause of death in children younger than 5 years –is responsible for approximately 81% of death in the first 2 years of life. ALRI is also known to be the leading cause of death in under-five children of Bangladesh where 55% of children aged 0-6 months are exclusively breastfed. Exclusive breast feeding (EBF) is considered to be a significant contributor to growth and reduced morbidity among children. However, the role of EBF on ALRI is still ambiguous, and the relationship between the duration of EBF and decreased risk of ALRI has not yet been thoroughly examined. Therefore, the objective of this analysis is to examine the effect of EBF duration on ALRI in first two years of life of the children living in a slum. Methods: We used data from the MAL-ED Bangladesh study. Duration of EBF status were categorized into seven groups (First is less than and equal to 30 days, second is 31 to 60 days, third is 61 to 90 days, fourth is 91 to 120 days, fifth is 121 to 150 days, sixth is 151-180 days and seventh group is more than 180 days). Incidence rate ratios (IRR) were calculated using generalized estimating equation (GEE) models with Poisson distribution. An exchangeable correlation structure was specified to account for within-child and within-family correlations of outcome measures. Associations were considered statistically significant for p- values less than 0.05. Results: After adjustment of potential confounders, we found that 31-60 days EBF children had 2.2 times higher risk, 61-90 days EBF children had 2.5 times higher risk, 91-120 days EBF children had 2.2 times higher risk, 121-150 days EBF children had 2.6 times higher risk and more than 180 days EBF children had 3.4 times higher risk of ALRI compared to children who were exclusively breastfed 151-180 days. Similarly male children had 1.6 times higher risk of ALRI compared to the female children. Autumn season had 1.4 times higher risk of ALRI compared to summer season. Conclusions: The findings of this study showed that EBF less than 151 days or more than 180 days, male children and autumn season were significant risk factors for ALRI in the first two years of life in Bangladesh slum children.

EM-030 ROLE OF SOCIO-ECONOMY ON PHYSICAL GROWTH OF ADOLESCENT DUNGRI BHILL BOYS Dr. Ashok Kumar Sahani Scientist, G.B Pant National Institute of Himalayan Environment & Development GRC, Srinagar, Garhwal, Uttarakhand-246174

The Dungri Bhil is a sub group of Bhill Tribe who were migrated to Gujarat from Rajasthan particularly from Mewar area of Rajasthan roughly about three hundred years ago. They were the subjects of Maharan Pratap Singh. When he was defeated by Muslim rulers, they dispersed to the hilly areas of Gujarat adjacent

166

to Mewar. Presently they are mainly distributed in the talukas of Meghraj, Bhiloda, Bijoynagar and Khedbrahma of Sabarkantha districts. The word Dungri derived from the word Dungar means hilly region/slops of hills where they clean the forest and used for settlement. After going through 12 successive statistical tables, it can be observe that the growth pattern of Dungri Bhill boys agree with the general growth pattern. The skin folds character shows a non-normal distribution at several ages. In the circumferential dimension i. e chest, upper arm and calf circumferences which give the composite measurement of bone, muscle and fat, the maximum annual increase has occurred in age group of 12+to13+ years and afterwards the width of the fat layer decrease during growth does not necessarily imply a decrease in the other dimension. The cross-sectional area of the fat stayed constant over a year during which the muscle and bone core enlarged, then inevitably the fat ring would be stretched out thinner. Still less does a decrease in fat layer width imply a loss of absolute total fat volume (Stuart & Reed, 1951) so because of the fat, it is seen that during successive ages the triceps and calf circumference measurement decreasing in the present data. The other traits of the current study show that the mean value in the proceeding ages increase accordingly. The velocity curve moving upward movement became peak at the age of 16+years in all measurement. The percentage of growth is also seen that at the age of 12+ & 13+ years the growth rate is higher in compare to 14+ & 15+ years. The slow process of growth again got its motion at the age of 16+years onwards. The dimensional measurement like bi-acromial and bi-illiac diameter reveals that at the initial age of adolescent the absolute growth capacity much higher in compared to other stages. The centre position of the stages remain in slow process and at the end i.e. at 16+ &17years again it came to growing process, but not like the percentage of growth like the age of 12+ & 13+years. The central percentile value of the measurement maintains almost balance. The present data on growth represent by mean value of 12 different body measurement displaying differential rhythm and time of peak velocities have been combined together to examine how for a pattern of growth occur in each stages and definite age of adolescent spurt can be trace out.

EM-032 Prevalence of congenital malformations in progenies from parental consanguinity within and outside Rajshahi City Corporation areas, Bangladesh M. Saiful Islam* and Sharmin Mustari Genetics and Molecular Biology Laboratory, Department of Zoology, University of Rajshahi, Rajshahi 6205, Bangladesh *Corresponding author; E-mail: [email protected]

Abstract: In a series of three separate studies during January through December 2018, various birth abnormalities (BA) from 178 clinical cases (Study 1); congenital malformations (CMF) from 70 consanguineous marriages (CM) and 100 non-consanguineous marriages (NCM) within the Rajshahi City Corporation (RCC) area (Study 2); and CMF from 48 CM and 12 NCM families outside the RCC area (Study 3) have been recorded, analyzed and interpreted. Results of Study 1 demonstrate that frequencies of miscarriages (5.69 vs. 3.00), stillbirths (3.85 vs. 1.00), postnatal deaths (8.97 vs. 1.00), preterm deliveries (29.49 vs. 38.00) and breech births (6.41 vs. 4.00) were significantly greater in CM compared to the NCM counterparts. In addition, both gestational ages and live birth weights of the children from CM were significantly lesser than those recorded from the NCM couples. Of 205 children derived from 70 CM couples in Study 2, 67 were suffering from various CMF, the remaining 138 children were normal. Compared to this, only 2 out of 267 children were affected from 100 NCM couples. Mental retardation (MR) represented the highest CMF (n= 37), followed by cerebral palsy (CP; n= 16), crossed-eyes (CE; n=4), blindness (BL) and microcephaly (MC; n= 3 each), deaf-mute (DM; n= 2), and Down’s syndrome (DS) and syndactyly (SD; n= 1 each). In contrast to the CM cases, the NCM couples had only two affected children, one crossed-eyed girl and the other polydactylous boy. The overall frequency of CMF in CM (32.68) was much higher than that in NCM cases (0.75). Of 103 children from 48 CM couples in Study 3, 38 (32.10%) were affected by various CMF, compared to only 2 affected children (4.00%) out of 25 from the unrelated NCM couples. In contrast to Studies 1 and 2, several new CMF, viz., abnormal voice (AV; n=5), cleft lip (CL; n=1), dwarf (DW; n=1) and non-articulated speech (NS; n=2), were recorded in Study 3. Moreover, the frequency of CMF in NCM couples (8%) in Study 3 was much higher than that found in Study 2 (0.75%), the variation might most probably be owing to the difference in the sample size. In relation to parental consanguinity, the importance of genetic counselling and pre-implantation screening relevant to the present findings has been emphasized in this report. Key words: Consanguineous marriages, birth abnormalities, congenital malformations, genetic counselling.

167

EM-033 Bias reduced and separation-free Poisson regression model for small or sparse count data Momenul Haque Mondol1; M. Shafiqur Rahman2; Wasimul Bari3 1Corresponding and presenting author's email: [email protected] 1Lecturer, Department of Statistics, University of Barishal, Barishal-8200, Bangladesh. 2Professor, ISRT, University of Dhaka, Dhaka-1000, Bangladesh 3Professor, Department of Statistics, University of Dhaka, Dhaka-1000, Bangladesh

Separation or monotone likelihood can be observed in the fitting process of Poisson regression using maximum likelihood estimation (MLE) technique when one or more parameters diverge to infinity. The separation is very common in count data when sample size is small or there is huge number of zero count or there is sufficient number of strong predictors or mixture of two or more such conditions. The study investigates the consequence of separation in the standard Poisson models and provides a solution by incorporating Firth's (1993) type penalty term, which was originally proposed for bias reduction in MLE, in likelihood score equation that we may call penalized Poisson. The modified score equation guaranteed an achievement of convergence and finite estimate of the regression coefficient. An extensive simulation study was conducted to assess the performance of penalized Poisson model over standard Poisson and Zero- inflated Poisson in the presence of separation. Several simulation scenarios were considered for creating complete or quasi-complete or near-to-quasi-complete separation by varying the sample size, proportion of event in the binary predictor which make separation, and the magnitude of regression co-efficient, log odds ratio β, relating the binary predictor to the response. The results revealed that the penalized Poisson having profile penalized likelihood based confidence interval performed better than the standard maximum likelihood based model, standard Poisson and Zero-inflated Poisson, in terms of bias, MSE and length of 95% confidence interval (precision) in all simulation scenarios. An illustration using real data also supported the simulation findings. Keywords: Separation problem; Bias reduction; Sparse count data; Zero-inflated Poisson; Poisson model.

EM-034 Level of serum high-sensitivity C-reactive protein predicts atherosclerosis and coronary artery disease in hyperglycemic patients Md. Saiful Islam1, Md. Rezaul Karim2, Kumkum Yeasmin3, Rowshanul Habib1 and Tanzima Yeasmin1* 1Department of Biochemistry and Molecular Biology, Rajshahi University, Rajshahi-6205, Bangladesh 2Department of Applied Nutrition and Food Technology, Islamic University, Kushtia-7003, Bangladesh 3Institute of Biological Sciences, Rajshahi University, Rajshahi-6205, Bangladesh *Corresponding email: [email protected]

Hyperglycemia is the essential characteristics of diabetes mellitus. Up to 30% of diabetic patients with myocardial ischemia remain asymptomatic and are associated with worse prognosis compared to non- diabetic counterpart, which warrants routine screening for CAD in hyperglycemic subjects. The aim of this study was to evaluate the clinical value of serum high-sensitivity C- reactive protein (hs-CRP) level in predicting the presence of coronary artery disease (CAD) & atherosclerosis in hyperglycemic subjects. A total of 201 study participants with 20-older years of age group were enrolled in this study and provided fasting blood samples. Subjects were selected from OPD (health checkup unit) of tertiary care hospital, Dhaka, Bangladesh. High-sensitivity C- reactive protein (hs-CRP) and Lipoprotein (a) were measured by particle-enhanced immunonephelometry according to the manufacturer’s protocol on a BN ProSpec analyzer. CAD risk of study participants was estimated with the use of Framingham Risk Score. There were total 91 healthy, 56 impaired fasting glucose and 54 diabetes mellitus study participants with a mean age range 44.30 ± 8.47, 47.0 ± 9.42 and 52.57± 11.94 years, respectively. Systolic blood pressures of the three groups were 114.91, 120.96 and 133.52 mmHg, respectively that were statistically significant with each other (p < 0.001). The average values of hs-CRP were 1.78, 3.28 and 5.23 ml/L, respectively. These increasing patterns were statistically significant. hs-CRP level was significantly (r = 0.295, p < 0.001) increases with the increasing concentration of fasting plasma glucose in blood. In terms of clinical application, high-sensitivity C - reactive protein (hs-CRP) is a stronger predictor of cardiovascular events (CAD & atherosclerosis) in hyperglycemic subjects and it adds prognostic information at all levels of calculated Framingham Risk. Keywords: High sensitivity C-reactive protein, Hyperglycemia, Coronary artery disease, Atherosclerosis.

168

EM-035 Correlation between malnutrition and morbidity pattern among children: A cross-sectional study Kumkum Yeasmin1, K M Saiful Islam2, Dewan Md. Mehedi Hasan3 and Tanzima Yeasmin4 1Institute of Biological Sciences, University of Rajshahi, Rajshahi - 6205. [email protected] 2Rajshahi Medical College Hospital, Rajshahi– 6000. [email protected] 3Directorate of General Health Services, Mohakhali, Dhaka – 1212. [email protected] 4Department of Biochemistry and Molecular Biology, University of Rajshahi, Rajshahi – 6205, [email protected]

Malnutrition, specifically underweight, has been considered as serious public-health problem that has been related closely with increased risk of mortality and morbidity of children. Females and children were vulnerable segment of population for suffering from diseases, associated with malnutrition.Therefore this study was carried out to assess the relationship between malnutrition and morbidity pattern among children in Rajshahi City, Bangladesh. This study was a community based, multistage cross sectional study, which carried out the period from January 2018 to December 2018. Height and weight were measured using standard procedure. Children whose measurements (z-score) fall below –2 SD of the reference population median were considered undernourished. A standard questionnaire was used for collecting the information about the morbidity pattern of the children.Chi-square test and multiple logistic regression analysis utilized in this study. The result shown thatTotal 1035 children were study population of this study. Among them, 538 were boys and 497 were girls. 33% of boys and 27% of girls suffered from diseases 1 – 3 times as well as 18% boys and 21.5% of girls suffered from various diseases 4 – 6 times in the previous three months of the survey. The study revealed that not stunted children (OR: 0.471; 95% CI: 0.338 – 0.657; p<0.01) had 0.471 times less chance to be attacked by diseases than stunted children. Not wasted children (OR: 0.20; 95% CI: 0.131 – 0.327; p<0.01) had 0.207 times less likely to be diseased in compared to wasted children. Whereas, not underweight children (OR: 0.62; 95% CI: 0.411 – 0.935; p<0.05) had 0.62 times less probabilities to be suffered from diseases than underweight childrenTherefore, the findings will be important for implementing of health and nutrition related programme, for the improvement of children’s health. Keywords: Disease, Malnutrition, Stunting, Wasting, Underweight

EM-036 An empirical study of Middle-income Farmers on Millet Production in Kano South Nigeria: A Stochastic Frontier Analysis Approach M. E. A. Khondaker1 M. U. Baba2 1Associate Professor, Dept. of Statistics, Hajee Mohammad Danesh Science & Technology University, Dinajpur, Bangladesh. Email: [email protected] 2PhD Research Fellow, Dept. of Statistics, HSTU & Dept. of Statistics School of Technology, Kano State Polytechnic, Kano State Nigeria. Email: [email protected]

This paper aimed at assessing the technical efficiency of millet farmers in Kano south Nigeria using the stochastic frontier production function from the parametric perspective. The determinants of technical efficiency, such as, education, extension access, access to loans, farm size, labour etc., which played a great role in technical efficiency level, are assessed to measure the level of efficiency of the middle-income farmers. 227 middle income farmers were selected for the study and provide an empirical analysis of the efficiency determinants which aim to find a way of improving the millet production and productivity. The results show that middle income farmers’ productivity is very weak ranging from 0.023t/ha to 5.62t/ha, averaging 1.07t/ha. While the technical efficiencies of middle-income farmers range from 0.013 to 0.930 with mean of 0.621. However, the results also indicate that most of the above determinants have negative effect on technical efficiency for the farmers. So, there is need by the government at all levels to review the existing agricultural policy that will favour the reality on the ground which will result in improving the output level of the middle-income farmers with their minimum available input. Keywords: Technical Efficiency, Middle income Famers, Productivity;

169

EM-037 Dimensionality Reduction of Histopathological Images for Breast Cancer Classification Using Deep-Transfer Learning Md. Shamim Reza1 and Jinwen Ma2 1 Pabna University of Science & Technology, Pabna, Bangladesh 2 Peking University, Beijing, China, [email protected]

Feature extraction plays an important role in building any intelligent system in the fields of statistical learning, pattern recognition and image processing. The performances of classification systems rely on appropriate data representation and therefore much of the efforts are dedicated to feature engineering step. The recent trend is to learn the features automatically with a deep learning framework like Convolutional Neural Network (CNN) which has been shown effective and state-of-the art in many computer vision tasks. More recently, transfer learning has been introduced via the CNN pre-trained model, which has been shown to be a potential feature extraction approach, being very fast and providing competitive results in many tasks. However, transfer learning base CNN pre-trained features resides high dimensional feature space and generally requires a large amount of memory, computation power, and wouldn't get desired results. Specially, in the application of histopathological image classification, high dimensionality could cause of over fitting. In this study, we deal the potential problem by proposing two adaptive feature transformation based on dimensionality reduction technique principal component analysis (PCA) and independent component analysis (ICA). To test the effectiveness of our methods, we firstly consider synthetic and general machine learning databases. On the basis of accuracy and different statistical inference including Friedman, BonferroniDunn test, we may conclude that our method reduces the features dimensionality and improves the classification performances considerably. Thereafter, we implement the methods on deep transfer learning features that have been obtained from Google InceptionNet and XceptionNet pre-trained models to diagnosis breast cancer from histopathological tumor images. In the classifying breast cancer disease, we consider BreakHist database obtained from P&D lab, Brazil. The database is composed of 7909 images captured from four magnification factor 40×, 100×, 200×, and 400× having two class benign and malignant tumor image. The image samples on the dataset are of size 700 × 400 pixel in RGB channel. For benign/malignant classification, our adaptive transformation method reduces the dimensionality as well as achieves improvement performance compared to directly extracted from pre-trained model, ranging from 1.5 to 4% accuracy at different zooming and provided state-of-the-art results for the seemingly more dificult 100×, 200× and 400× zooming level. The empirical findings of this study provided a better understanding of neural feature extraction from pre-trained model and one more steps to transfer learning for computer-aided diagnosis in the field of medical image application. Keywords: Dimensionality reduction, Feature extraction, CNN, Machine learning, Transfer learning

EM-038 Improved Machine Learning Model for Diabetes Types Classification by Using Integrated Data Balancing and Features Elimination Techniques Afsana Afroz, Aliul Islam, Sirazum Monira, Woomme Kulsum, Shamim Reza Department of Statistics, Pabna University of Science & Technology, Pabna, Bangladesh [email protected]

Background and objective: Diabetes stands among the top 10 causes of death and poses as an alarming health issue worldwide. It is a common chronic diseases which causes an increase in blood sugar. Early diagnostics and treatment can increase the chances of survival. The rise in machine learning (ML) approaches improve the diagnostic process much faster and increase the consistency of pathologists. Though for the last few years ML have shown promising results on classification or recognition task but it poses a significant challenge when the dataset is class imbalanced or skewed class distribution. Most of the ML algorithms assume a balanced distribution or the same distribution of classes for training. In imbalanced dataset, there are a large number of majority class instance, a classification algorithm tends to favor majority classes while there is a chance to incorrectly classifying the sample from the minority classes. Moreover, as the medical datasets are not suficiently large and imbalanced among the classes, the direct training often leads to the overfitting problem, i.e. being inability to generalize the knowledge. Data: The Pima Indian Diabetes dataset obtained from UCI repository of machine learning was utilized for this study. The dataset is comprised of 768 patients, of which 268 tested positive and 500 tested negative diabetes. We have also used a more practical dataset collected from a known population. In collaboration with the Pabna diabetes hospital, Pabna city, we extracted information from patient records who were tested

170

for diabetes at the medical facility. This dataset consists of 465 patients, 10 attributes, having 373 diabetic positive and 92 free from diabetic. Both of the datasets shows heavily imbalanced between the class. In the PIMA diabetes data, the diabetes negative class is represented 2 times as much as positive class whereas Pabna diabetes hospital data, negative class 4 times as positive class. Methods: The motive of this study is to reduce the bias towards the plentiful classes, preventing over fitting in ML model and improving the performances. To address the imbalance issue, we have considered under- sampling, over-sampling and synthetic minority over-sampling technique (SMOTE) in ML training. In order to reduce irrelevant features, logistic regression have used to eliminate the attributes of which P value less than 0.05. Protocols and results: As experimental protocols to evaluate the results, the dataset has been randomly divided into a 70% training, and 30% testing set by using stratified random sampling. We have also conducted five-fold cross validation to examine the models. The results show that the imbalanced data has significant impact on the performance of ML model. ML with a balanced and significant feature lead to better composition of precision, recall, f-score as well as better classification accuracy compared to ML trained with imbalanced training data, ranging from 0.7 to 6%. The proposed method can also be extended to medical datasets other than diabetes, so that these types of deadly deceases can be correctly diagnosed for the betterment of humanity. Keywords: Diabetes, Machine learning, Imbalanced data, SMOTE, Oversampling, Logistic regression

EM-039 Impact of Air Pollution on Street Hawkers Nusrat Karim Turna and A H M Saadat Department of Environmental Sciences, Jahangirnagar University, Savar, Dhaka-1342 Corresponding Author: [email protected]

Main cause for which Bangladesh stands second lowest position in EPI 2018 among 180 is poor air quality and street hawkers are the worst sufferer of it. A structured survey was conducted among 146 street hawkers in Mirpur, Dhaka to assess the health impact of air pollution. Among the respondents 79 % were male and 21% were female with age ranges from 14 to 55 years. 55% of the hawkers were illiterate and came from lower middle-class family. Duration of working hour varies from 4 to 14 hours where most of them works up to 12 hours. Their income starts from 8000 to 55000 taka where most of them earns 10000-20000Tk. Survey suggests that 75% of street hawkers were affected by several diseases like asthma, tuberculosis, skin disease, irritation, cough etc. due to air pollution. Male hawkers are more affected by asthma and female hawkers are more affected by cough. Only a few of them use personal protective equipment and due to low literacy lack of awareness about air pollution is common in them. Study shows that most of them have weaker lung in comparison to healthy people. They don’t get enough opportunity of medical services due the nature of their work. Appropriate policy and action required to ensure health working condition of this deprived group of people. Keywords- Survey, respiratory health hazards, air quality, street hawkers, lung strength, Dhaka city.

EM-040 The evolution of present livelihood of DHULI: A Statistical overview on North-western part of Bangladesh Hossain Ahmed Tanvir*, Tanveer Al Azad**, Nowrin Ahmed*** * Final year student, Department of Geology and Mining, University of Rajshahi. Mail: [email protected] ** 3rd year student, Department of Journalism and Mass communication, University of Rajshahi. Mail: [email protected] *** MFA in Oriental Arts, Department of painting, Oriental Art & Printmaking, University of Rajshahi. Mail:[email protected]

Dhol is an ancient beat instrument with cylindrical to spherical shaped wooden body covered by animal skin on both open sides and played by bamboo sticks and hand fingers. History of its origin in ancient past can be traced where it was used in various traditional festivals like “puja, musical orchestra, jatra pala, stage drama, various announcement and welcoming the guest” in this Indian subcontinent and Bangladesh. Now- a-days this tradition is transformed by the impact of globalization like playing records on loudspeaker or amplifier which replaced the sound of dhol and pressed an adverse effect on the life and livelihood of DHULII. Previous studies show various result on evolution of folk music and folk cultures but this article mainly focused on the present livelihood of DHULI from multi-dimensional disciplinary approach (science, social science and arts). Both qualitative and quantitative research methodology are used to address the

171

observation on this regarding issue (age 30 , both male and female ) over the area for interviewing about their present and previous life and livelihood as DHULI professionals by using non probability sampling. The findings are the statistical≥ 푦푒푎푟푠 observation표푙푑 about the impact of globalization on present livelihood condition of DHULI.

EM-041 Performance of different data mining methods for predicting rainfall of Rajshahi district, Bangladesh Md. Mostafizur Rahman * and M. Sayedur Rahman Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh, Corresponding E-mail (*): [email protected]

Rainfall forecasting has been treated as one of the most challenging task around the world. Accurate and timely rainfall prediction can be very useful to take necessary safety action for different purposes such as ongoing construction projects, transportation activities, agricultural tasks, flight operations and flood situation, etc. Due to climate variations, an accurate rainfall prediction has become more complex than before. Data mining techniques can effectively predict the rainfall by extracting the hidden patterns among available features of past weather data. This paper contributes by exploring the use of various data mining techniques such as Classification and Regression Trees (CART), Support Vector Machine (SVM), K- Nearest Neighbors (K-NN), Random Forest (RF) and Linear Discriminant Analysis (LDA) for rainfall prediction in Rajshahi city, Bangladesh. For effective prediction, pre-processing technique is used which consists of cleaning and normalization processes. Performance of used data mining techniques is analyzed in terms of precision, recall and f-measure and overall accuracy and different graphical comparison tools. Keywards: Classification and Regression Trees, Support Vector Machine, K-Nearest Neighbors, Random Forest, Linear Discriminant Analysis, Rainfall.

EM-042 Comparison among selective NGOs in Bangladesh: A case study from Rajshahi and Natore district Md. Alamgir Hossain*, Md. Mostafizur Rahman, Md. Nurul Islam and Md. Monsur Rahman Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh, Corresponding E-mail (*): [email protected]

The established expression of opposition between state and NGOs as development agents has shifted to one of complementarities and common interest. Along with this, the ‘comparative advantage’ claimed for NGOs has expanded from economic and welfare benefits to encompass also the political goods of civil society and popular participation. This paper reviews these developments in the context of Bangladesh. It argues that they need to be assessed critically in ways which are both theoretically informed and locally contextualized. While recognizing that there are, indeed, areas of common experience and interest between the state and NGOs in Bangladesh, it questions whether these necessarily coincide with the interests of those they all invoke: the poor. Keywords: Micro-credit program, comparison index, Grameen Bank (GB), BRAC, ASA, JCF and URN.

EM-043 Evaluation of Facebook Addiction and its Associated Factors among Varendra University Students: A Cross Sectional Survey Study Ahammad Hossin1, Rasel Mahamud1, Md. Rasel Bishwas1 & Dr. Md. Kamruzzaman2 1Assistant Professor, Dept. of CSE, Varendra University, Email: [email protected] 1B.Sc. (Hon’s) student, Dept. of CSE, Varendra University, Email: [email protected] 1B.Sc. (Hon’s) student, Dept. of CSE, Varendra University, Email: [email protected] 2Associate Professor, Institute of Bangladesh Studies (IBS), University of Rajshahi, Email: [email protected]

Now a day’s Facebook addiction of university students is an immense concern over the world. The main objective of this study was to detect the evaluation of Facebook addiction and its associated factors among the students of Varendra University (VU). This study was conducted with primary data collected from face to face questionnaire survey through simple random sampling without replacement (SRSWR) of 326 students (sample size) from 5000 students (population size) under the 10 departments of Varendra University. The statistical analysis like frequency distribution, numerical statistics, and Binary Logistic Regression (BLR) analysis were conducted using IBM SPSS of version 25. This study reveals that majority of the students of VU follows 51.2% students were not addicted on Facebook and 48.8% were addicted on

172

Facebook of which a little bit addiction was 30.4%, moderate addiction was 10.4% , severe addiction was 5.5 and extremely sever addiction was 2.5%. BLR model of Facebook addiction on categorical covariates was fitted. The significant odds were found for covariates- marital status, using Facebook everyday, use Facebook after going to bed and use of Facebook of previous night of exam, respectively. The proposed BLR model shows 67.8% overall accuracy of prediction of Facebook addiction. Keyword: Facebook addiction, university students, odds ratio, BLR model and prediction.

EM-044 Insulin secretory defects and determinants of attending at a tertiary hospital in northern region of Bangladesh K Hasnin, R karim, L Ali Rajshahi Diabetic Center, Rajshahi; 2Department of Biochemistry and Cell Biology, Bangladesh University of Health sciences (BUHS), Dhaka, Bangladesh; E mail: [email protected],[email protected], [email protected]

Introduction: Rational management and prevention strategies for DM require an understanding on the burden, determinants and basic defects of the disorder which vary substantially among populations due to racial and environmental heterogeneity. Limited information on the prevalence, risk factors and basic defects of DM are present among Bangali population in general but, so far, no study has yet been conducted on any epidemiological or pathophysiological aspects of DM in Rajshahi population. Objective: This study was conducted to characterize the new patients attending the Rajshahi Diabetic Center which gives an idea about the proportion and risk indicators of DM among people in the Rajshahi region. At the same time the study also aimed to investigate the basic defects of T2 DM in this population. Methods: It was a hospital based observational analytic study with a hybrid research strategy having both cross-sectional and nested case-control designs. A multistage sampling technique was followed with 660 subjects in stage 1 for exploring the proportion of T2 DM and its association with clinical, socio- demographic and biochemical risk factors. In stage 2, a subgroup of subjects and non-diabetic controls, selected through a nested case-control design, were investigated for the basic defects (insulin secretory defect or insulin resistance) of DM and their association with various risk factors. Nutritional intake was assessed by food frequency questionnaire (24 hr dietary recall) method. Results: Out of the total subjects attending the OPD of RDC for the first time, 65% were found to suffer from T2 DM. The diabetic subjects were mostly (75%) of middle age (37% within 30-40 yrs and 37% within 41-50 yrs), only 2% cases had age <30 yrs and 23% had age >51yrs. There was a male preponderance in the proportion of T2 DM (56.4% female and 76.7% male). The diabetic subjects were mostly from middle (34.7%) to upper middle class (31.6%) background and by occupation females were mostly housewives (77.7%). Out of the diabetic subjects 37.7% had lower educational levels. The mean ±SD calorie consumption (2549±637) of the diabetic subjects was higher than that in non-T2 DM subjects and it came mainly from CHO (59.3%) and fat (55.9%). All the newly diagnosed diabetic subjects in RDC were found to have poor diabetic control when judged by their blood glucose levels. Lipid levels were also found to be largely uncontrolled in the diabetic subjects. On multiple regression, FBG was found to be associated with age (p=0.002) and occupation (p=0.026) in females and age (p=<0.001) and TG level (p=0.026) in males T2 DM subjects. Fasting serum Insulin was significantly higher in the T2 DM group as compared to non-T2DM (Serum Insulin level µIU/ml, M±SD) (13.5±4.9) subjects (p=<0.001). HOMA%B was (37±17) significantly lower in the T2 DM subjects as compared to non-T2 DM subjects. HOMA%S was (41±13) significantly lower in the T2 DM subjects as compared to non-T2 DM subjects. On multiple regression analysis with HOMA%B was found to be significantly associated with sex (<0.001) and education (p=0.010). On similar analysis, with HOMA%S was found to have significant association with males (p=0.002) and daily CHO intake (p=0.028). Conclusions: It may be concluded that both (HOMA%B) and (IR) constitute the basic defects of diabetes in Rajshahi population, but (HOMA%B) seems to be more predominant in these subjects. (HOMA%B) in Rajshahi population is associated with males and level of education and insulin resistance (IR) is associated with males and daily CHO intake.

EM-045 The use of Polygenic Risk Scores as an Instrumental Variable in Mendelian Randomization Studies: A Systematic Review Shabab Noor Islama and Md. Jamal Uddina a Department of Statistics, Shahjalal University of Science & Technology, Sylhet-3100, Bangladesh.

173

EM-046 Statistical analysis and application of machine learning models for prediction and prognosis in Gastric cancer studies Md. Shadikug Jaman1, Nishith Kumar2, Md. Aminul Hoque3* 1 Department of Statistics, University of Rajshahi, Rajshahi, Rajshahi Bangladesh 2 Department of Statistics, Bangabandhu Sheikh Mujubur Rahman Science and Technology University, Gopalganj, Bangladesh 3 Department of Statistics, University of Rajshahi, Rajshahi, Rajshahi-6205, Bangladesh * E-mail: [email protected]

Illness and health problem may be the major and high concern not only in Bangladesh but also throughout the world due to known and unknown communicable and non communicable diseases. Every country of the world is concerned about health problem because the total development of a country mainly depends on health. Gastric cancer is one of the major health hazards and the second leading cause of death from malignant disease worldwide. This is the common cause of cancer-related death in human population. However this killer disease should be addressed properly to accurate prediction and prognosis to prevent the health hazard and to attain one of the important SDG goals. The accurate prediction of survival rate in patients with gastric cancer remains a challenge due to the increasing complexity of cancer, treatment protocols, and various patient population samples. To achieve this goal we have used statistical analysis and machine learning models in this research study. Aside from data normalization and classification, the applied machine learning methods provide promising results in terms of accuracy of survival time prediction and prognosis of gastric cancer. Throughout the analysis of our study we used machine learning (ML) models such as the linear Support Vector Regression, Lasso regression, Kernel Ridge regression, K-neighborhood regression, Kernel SVD, BN analysis and Decision Tree regression etc. These models achieved most accurate survival prognosis results. We have compared the results under different machine learning models with gastric clinical Japanese data and found that Lasso regression and Kernel SVD showed the better performance. Keywords: Gastric Cancer, Prediction, Prognosis, Machine learning, Lasso Regression, Kernel SVD.

EM-047 The Genesis and Journey of Silk Economy and its Past, Present and Future. Md. Abdul Hakim Director General (DG), Bangladesh Sericulture Development Board

Silk is a high-tech industry. It is half agricultural and half industrial. Its demand as well as production is ever increasing. Its production fields and export destinations are undergoing dramatic changes. It has textile as well as non textile aspects. Its full potentials are yet to untap. Its multidimensional researches and utilizations may bring about a qualitative change in human civilization and lifestyle. This paper is a humble attempt to portray and analyze its past, present and future along with its huge potentials from economic and commercial point of view. About two thousand years ago during the regime of Han dynasty in China the Silk Road was a network of trade routes of economic, cultural, political, and religious interactions .The Chinese made the Great Wall for the safety and protection of their trade products and routes. The Silk Road trade had a key role in the development of the civilizations of China, Korea, Japan, India, Iran, Europe, Africa and Arabia. Not only silk but also many other goods, philosophies, sciences, technologies and ideas were also exchanged in this corridor. Cotton, silk and wool are only three natural fibers which constitute the major part of basic need textile industries connected to many macro economic indicators and SDGs. Cotton is the king of textile, silk is the queen and wool is their daughter. Silk economy unlike any other economy has some distinctive features. It is labor intensive and poverty alleviator. Rate of labor participation in this industry is very high. Mechanization and labor cost affects the production cost of raw silk. High level skill is a factor and continuous training is a prerequisite. The poor are producers and the rich are consumers. Only a few country are in production and the rest are in consumption. It is a cash crop and a high value agriculture. It is multiple cropped and cottage in nature. Even the family children can earn some money through this cropping. Its

174

impact and outcome are heavy and diversified. Examples are ural development, women empowerment, easy employment scope, environment protection , industrial expansion, economic disparity remover, facilitators of the fair distribution of wealth etc. But it is very difficult to make it commercially and economically viable.

EM-048 Identification of Outliers in Gene Expression Data A.H.M. Rahmatullah Imon Department of Mathematical Sciences, Ball State University, Muncie, IN 47306, USA E-mail:[email protected]

Identification of outliers is a big challenge in big data although it has drawn a great deal of attention in recent years. Among all big data problems detection of outliers in gene expression data warrants extra attention because of its inherent complexity. Although a variety of outlier detection methods is available in the literature Tomlins et al. (2005) argued that traditional analytical methods, for example, a two-sample t- statistic, which search for common activation of genes across a class of cancer samples, will fail to detect cancer genes which show differential expression in a subset of cancer samples or cancer outliers. They developed the cancer outlier profile analysis (COPA) method to detect cancer genes and outliers. Inspired by the COPA statistic, some authors have proposed other methods for detecting cancer-related genes with cancer outlier profiles in the framework of multiple testing [see Tibshirani and Hastie (2007), Wu (2007), Lian (2008), Wang (2010)]. Such cancer outlier analyses are affected from many problems specially if there is any outlier in the data set then classical measures of location and scale are seriously affected. So the test statistic using these parameters might not be appropriate to detect outliers. In this study, we try to robustify some existing methods. We propose three new techniques Expressed robust t-statistic (ERT), Modified Outlier robust t-statistic (MORT) and Least Sum Square of Ordered Subset Robust t-statistic (LSOSRT) for the identification of outliers. The usefulness of the proposed methods is then Carlo simulation and real cancer data. Keywords: Gene expression, Cancer outlier profile, Multiple outlier, Masking, Swamping, Robust statistics.

EM-049 Determinants of Ever Injury of Street Children of Dhaka City in Bangladesh: An Approach of Logistic Model Md. Shahiduzzaman1, M. Korban Ali2, Md. Nurul Islam3, Md. Rafiqul Islam4 1Assistant Professor, Department of Electrical and Electronic Engineering, Northern University Bangladesh, E-mail: [email protected] 2Professor, Department of Computer Science and Engineering, Manarat International University, Bangladesh, E-mail: [email protected] 3Professor, Department of Statistics, Rajshahi University, Bangladesh, E- mail: [email protected] 4Professor, Department of Population Science and Human Resource Development, Rajshahi University, Bangladesh, E- mail: [email protected]

Health condition of street children is an important matter for the socio economic development of a country. Injury is a common problem of street children in the country like Bangladesh. So it is important to know the factors which are affecting the ever injury of the street children for the interest of the government agencies and policy makers of Bangladesh. A sample of 1108 data were collected from street children in different places of Dhaka City using a comprehensive questionnaire. A univarite and a bivarite tables were used to 2 analyze the acquired data. The analysis was also made using percentages and proportions. χ - statistic was used to find the association between the ever injury and background characteristics of the street children. A logistic regression analysis was also utilized to find the determinants of ever injury among the street 2 children. In χ - statistic, it was found that types of street children, sex of street children, place of interview, reason of becoming street children, current age, earning per day, saving money, parents composition, food collection, smoking practice and drug addiction were statistically significant with ever injury of street children. In logistic analysis, it was observed that place of interview, current age, saving money, parents composition, food collection and smoking practice were the significant predictors for ever injury of the street children. Considering the above findings it can be suggested that the government should take necessary steps so that the street children avoid their staying from the vulnerable places like mazar, railway station/bus stand/launch terminal etc. and to control the movement of street children to reduce the ever injury with increasing of their age. Along with these it can also be suggested to improve social relation between the

175

partners of the parents to reduce divorces and parents should provide food to their children. A multidimensional measure may be undertaken to reduce the ever injury due to smoking. Keywords: Street Children, Ever Injury, χ2-test, Logistic Regression, Dhaka City

EM-050 Development of Social Communication System- SEENEVER for the Visually Impaired People Md. Tahmid Shariar Khan1, Pritom Sarker1, Amit Chakraborty1, Julker Nayem1, Nakib Aman Turzo2, Ahammad Hossain2, Md. Aminul Islam3 & Md. Ayub Ali5 1B.Sc. (Hon’s) Student, Dept. of CSE, Varendra University, Email: [email protected] 1B.Sc. (Hon’s) Student, Dept. of CSE, Varendra University, Email: [email protected] 1B.Sc. (Hon’s) Student, Dept. of CSE, Varendra University, Email: [email protected] 1B.Sc. (Hon’s) Student, Dept. of CSE, Varendra University, Email: [email protected] 2Lecturer, Dept. of CSE, Varendra University, Email: [email protected] 2Assistant Professor, Dept. of CSE, Varendra University, Email: [email protected] 3Adjunct Faculty, Dept. of MSJ, ULAB, Email: [email protected] 4 Professor and Chairman, Dept. of Statistics, University of Rajshahi, Email: [email protected]

Social networking has become a global phenomenon and being one of the buzzword by means of communication. The internet has appeared as the most effective means of disseminating information. It is worth mentioning that social networking systems like Facebook, Twitter, Tumblr, Reddit, Linkedin, Pinterest, Myspace etc. have been in use since its launching. These social networking systems have emerged a popular social networking site used by people of different ages and professions for various purposes. No sole social platform has been created with the primary focus of achieving accessibility to the visually impaired people. The main objective of this project was to develop social networking system for visually impaired people. We named it Seenever. Seenever is, in plain words, a social communication website for the visually impaired (partial or fully) individuals with easy voice control, gesture, haptic feedback & amp; matrix driven layout. It offers an intuitive yet simple user interface and command reconfirmation protocol that can be controlled without needing any extra hardware device. Moreover, it inherits all the features from modern online social platforms. Current modern social platforms lacks the accessibility options for low/zero eyesight, so users are being forced to use third party software plugins or expensive devices to assist them browse and use those social sites. Hence, we felt the necessity or creating a social site with accessibility for visually impaired audience in mind. Primarily focused on accessibility, the website control is made to be usable for handheld devices and desktop/laptop computers using 4 techniques, evenly proportioned matrix like layout and multi-directional linear data flow webpages, haptic feedback, accessible distinct keyboard keys, mouse & amp; voice control. Inputs are driven by speech recognition and feedback to reconfirm and modify if it was the desired input user wanted. There are multiple ways to do each task, so users can adapt to their preferences. It also can be used in an ordinary way that feels natural even for those who can see clearly and have less sight problem. The website comes with two themes, one of high contrast and a default visually pleasing one for normal users. Keywords: Seenever, Social Communication System, and Visually Impaired People

EM-051 Shifting Season in Bangladesh: An Alarming to the Farmers Md. Ashek Al Naim1, Md. Abeed Hossain Chowdhury2, Md. Ayub Ali3 1 Research Student, Department of Statistics, University of Rajshahi, Email: [email protected] 2 Project Manager, BARC, Farmgate, Dhaka, Email: [email protected] 3 Professor, Department of statistics, University of Rajshahi, Rajshahi 6205 E-mail: [email protected]

Recently the farmers are facing some problems in cultivating rice like AUS, AMAN and BORO due to irregular rainfall. The present investigation is about whether there is any shifting of season or not. Rainfall, Maximum temperature and Minimum temperature data are used here. Method of pattern matching is considered to see the shifting of the season. Compare to BARC (April to July for AUS, June to November for AMAN, November to April for Boro) February to May for cultivating AUS, May to October for cultivating AMAN and October to February for cultivating BORO was proposed here in this study. Based on this finding, farmers will be benefitted and government should take proper steps to aware farmers and help in increasing rice production. Keywords: Season shifting, Rainfall, Maximum Temperature, Minimum Temperature, Pattern matching, BARC

176

EM-052 Risk factors for diabetes and drug utilization pattern among the diabetic patients Tripti Rani Paul1*, Sirajum Monika1, Hasan Hafizur Rahman1, Mir Imam Ibne Waheed2 1Department of Pharmacy, Faculty of Science and Engineering, Varendra University, Rajshahi-6204, Bangladesh. 2Department of Pharmacy, Faculty of Science, University of Rajshahi, Rajshahi,-6205, Bangladesh. Correspondence Email: [email protected]

Diabetes, the seventh leading attributable risk factor for burden of disease in South Asian region and most of the adults with diabetes live in low and middle-income countries. This prospective cross sectional study was designed to analyze the risk factors for diabetes and also the drug utilization pattern among the patients. This survey was conducted in different diabetic centre of Rajshahi city situated in north-west part of Bangladesh. About 500 prescriptions were collected from diabetic patients for a period of 12 months from March 2018 to April 2019. For this purpose, a self-designed standard questionnaire was developed and all information was recorded from the prescriptions and directly interviewing the patients. The prescribed drugs were classified according to WHO Anatomical Therapeutic Chemical Classification System. Data were analyzed using statistical software SPSS version 20. A total of 500 prescriptions were analyzed of which 38.46% male and 61.60% female among them 49.4% patients were housewife. Maximum number of patients, 28.2% belongs to the age group 41-50 years, beside 4.8% and 21.2% patients were in age groups 21-30 and 31-40 years respectively. Among the patients 92.2% were suffering from type 2 diabetes, 4.4% of type-I and 2% of gestational diabetes, whereas 26.4% diabetic patients were unaware of physical exercise. Patients with diabetes had risk factors for family history of diabetes (30%) and hypertension (45.6%). Other risk factors include overweight (38.2%), smoking (20%), obese (9.8%) and alcoholic (1.4%). Drug management of diabetes include insulin (42.6%), oral hypoglycemic agents (76%) and/or both insulin and oral medicine (26.2%). Among the prescribed oral medicine 57.8% patients were treated single drug, 17.6% with two drug combinations, however three drug combinations were rarely used (0.6%). In case of monotherapy, metformin HCl (24%) and gliclazide (13.2%) were preferred, in some cases gliclazide and metformin HCl combination (6.2%) were also advised. This study revealed that people are suffering from diabetes irrespective of age and multiple modifiable risk factors were involved in the pathogenesis of diabetes.

EM-053 A First Longitudinal Growth Study in Jessore District Md. Shahjada Ali Nawaz1 and Md. Ayub Ali2 1 PhD Fellow, Department of Statistics, University of Rajshahi, Rajshahi 6205 E-mail: [email protected] 2 Professor, Department of Statistics, University of Rajshahi, Rajshahi 6205 Email:[email protected] Corresponding Email:[email protected]

Longitudinal growth study helps us to understand the growth status of the children through the growth chart and growth models on longitudinal morphological data. There were 135 primary schools in Jessore Sadar Upazila and they are divided into Urban School and Rural School. Urban area has 35 primary schools and rural area has 100 schools. Three primary schools from Urban and four primary schools from Rural are chosen at random. Data was collected longitudinally from November 2006 to November 2009. Children from class I in the year 2006 of every selected school were considered as the sample observations. To the best of our knowledge, this is the first time that morphological data was collected longitudinally from 6-18 years aged student of Jessore district. Several growth charts e.g., height for age, weight for age, weight for height, sitting height for age and so on, were constructed. Medical practitioners as well as the growth scientists might use this kind of charts. Keywords: Longitudinal growth chart, height for age, weight for age, weight for height

EM-054 Evaluation of Drynaria quercifolia extract for in vitro anticholinesterase and antioxidant activity relevant to treatment of Alzheimer’s disease. Rafia Ferdous, Md. Golam Sadik Department of Pharmacy, University of Rajshahi, Rajshahi-6205, Bangladesh.

Alzheimer’s disease (AD) is a progressively developing irreversible neurodegenerative disorder in the elderly. The symptoms include memory loss, aphasia, mood swings and behavioral abnormalities. Recent

177

estimates indicate that AD is a third leading cause of death just after cancer and heart disease. Cholinergic dysfunction and oxidative stress have been implicated as the major contributors to the pathogenesis of AD. Therefore, inhibition of cholinesterase and oxidation offer the two promising strategies in treating AD. Three drugs currently approved by the FDA to treat AD can ameliorate the symptoms; however, none of them can completely restrict or reverse the progression of AD. Therefore, there is still an urgent need for pharmacologically safe, cost effective and immediately available agents with minimal side effects. Traditional medicines prepared from plants and herbs have been found useful in the treatment of AD. Drynaria quercifolia, known as Pankhiraj, in Bengali, belongs to the family Drynariaceae and is distributed all over Bangladesh on the tree trunk and shady old walls. It is a very common lithophytic fern with fleshy, creeping woody rhizome with brown scales and dimorphic fronds. In Bangladesh, the plant is traditionally used in the treatment of mental disorder. In addition, the plant is also traditionally useful in long sleeping disorder, insanity, phthisis, hectic fever, dyspepsia, skin diseases, helminthes, cough, poultice to swellings, typhoid, jaundice, dysentery, scanty urination, spermatorrhoea, rheumatic pain and so one. Therefore, the objective of the study was to explore new compounds related to potential anti-AD activity from Drynaria quercifolia that may provide a lead for the treatment of AD. The study began in the month of August 2019 with collection of plant materials, preparation of sample and crude extraction in methanol followed by assay of total phenol, flavonoid and antioxidant activity against DPPH, Ferric radical, etc. The crude extract demonstrated appreciable phenolic and flavonoid content and a notable antioxidant power. It also demonstrated butyryl cholinesterase inhibitory activity. Further study continues to bioassay guided fractionation of the crude extract and purification of anti AD related compounds.

EM-055 Geographical Efficiency of Daily Calorie Intake in Rajshahi District of Bangladesh Most. Mira Khatun1* and Provash Kumar Karmokar Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh *Communicating Author: [email protected]

Rajshahi is an important agricultural crop production area in Bangladesh. Most of the people of this region depend on the agriculture for food calories and their livelihoods. The attainment of the nutritional levels may differ in different locations for different causes. Getting nutrition of the people in this territory is currently going through a bit crisis, and verifying whether the level of achievement of nutrition can recover part of its capability is now a priority. As therefore attention is now focused on estimating whether the households can improve their level more efficient due to geographical regions. This paper aims to estimate geographical technical efficiency in the households of Rajshahi districts following stochastic frontier modeling approach. In order to assess whether geographical impact exists in the technical improvements in nutritional level or not a set of 350 households data were collected following a well-structured questionnaire. The stochastic production frontier is designed to estimate the underlying efficiencies used in this research. The mean geographical efficiency is found to be 74.5% and the inefficiency is inherently unobservable, being convoluted with traditional stochastic shocks has been noticed in this study. The daily calorie intake has calculated with the help of the daily food consumption for the selected geographical locations and classification have been made to make a comparison among such groups. Boalia and Paba were identified as more secured territory while the Bagha as lowest efficient location of performance among the selected thanas. The joint effect of inefficient variables CF, NF and Credit Access (Yes) were revealed by the Likelihood Ratio Test (LRT) from stochastic frontier model. Finally 79% of the mean efficiency for the selected five geographical locations and together with the consigned inefficient variables has been confined. As therefore our findings have firm implications for what policies should be implemented in order to improve efficiency in nutritional aspects and that could be contributory in concern of the SDG implementation goals. Keywords: Error Components Frontier, Technical efficiency, Geographical Locations, Daily Calorie Intake.

EM-056 Multivariate Correspondence Analysis in Patterning Tobacco Use in Bangladesh Suraiya Shikha1 and Papia Sultana2 1 Department of Statistics, University of Rajshahi, Bangladesh. E-mail: [email protected] 2 Department of Statistics, University of Rajshahi, Bangladesh. E-mail: [email protected]

Bangladesh is a low-income country and one of the largest tobacco consuming countries in the world, smoking causes about 25% of all deaths in Bangladeshi men aged 25 to 69 years and an average loss of seven years of life per smoker in Bangladesh. This study aimed to reveal pattern of tobacco use in

178

Bangladesh and identify their correlates using multivariate correspondence analysis (MCA). This study was based on the secondary data collected by the Global Adult Tobacco Survey (GATS), Bangladesh, 2009–10. It had been found that among the respondent who were not using any form of tobacco product 68.12% female and 31.88% male. Whereas, among the respondent who were smoking tobacco 2.64% female and 97.36% male; 69.46% were female and 30.54% were male among those who were taking smokeless tobacco; and among both smoking and smokeless tobacco user 5.14% were female and 94.86% were male. From multivariate correspondence analysis we found that people of age 35 to 54 years, less than primary completed, unsecured job, with low household wealth index etc. were corresponded to tobacco smoking and both (smoking and smokeless) tobacco use. On the other hand, people with age 55 years and above, had no formal schooling, home maker, with lowest household wealth index were corresponded to smokeless tobacco use. Therefore, policy makers and government should active the quitting tobacco use campaign more for disadvantaged people. Keywords: Smoking pattern, Correspondence analysis, correlates, GATS. Submission for Student Paper/Poster Competition? Yes

EM-057 Inhibition of carrageenan induced mice paw edema and stabilization of HRBC membrane by isolated compounds and fractions of Magnolia champaca (L.) stem bark Md. Mahadi Hasan, Masuma akter, Md. Ekramul Islam, Mst. Shahnaj Parvin* Department of Pharmacy, University of Rajshahi, Rajshahi-6205, Bangladesh *Correspondence: [email protected]

Magnolia champaca L. is documented in Asian country as a traditional medicinally necessary plant. So as to scientifically evaluate a number of the folkloric and ethnomedical uses of this plant, the aim of this study was to explore the anti-inflammatory and antineoplastic properties of the stem bark of the plant. In vitro anti- inflammatory assay was performed by human red blood cell (HRBC) membrane stabilization assay. In vivo anti-inflammatory activity of the isolated compounds was investigated by carrageenan induced paw edema & acetic acid induced vascular permeability assay in mice. In in vitro anti-inflammatory assay CHF was highly effective in inhibiting heat & hypotonicity induced hemolysis of RBC with the values of 73.91%, 75.92% as compared with standard acetylsalicylic acid which showed 84.34% & 83.32% protection at 500 μg/ml concentration. Pure compound Trans-Syringin (C-1) and an impure one (C-2) were isolated from CHF using conventional column chromatography and the structure of the isolates were elucidated by 1H and 13C NMR data as well as by comparison with reference data. In in vivo anti-inflammatory assay a dose dependent and statistically significant (p<0.05) decrease in paw edema, leukocytes migration and vascular permeability were observed in the treated groups. Mice treated with C-1 & C-2 at 50 mg/kg bw (IP) dose exhibited 60.26% & 56.30% inhibition in paw edema formation at 6 hr post carrageenan administration; 61.89% & 50.13% in leukocytes migration and 70.24% & 59.52% in vascular permeability, respectively. So, the present study revealed that the isolated compounds and fractions of stem bark of Magnolia champaca possess anti-inflammatory property which support its use in traditional medicine and suggesting that the plant may be further investigated to discover its pharmacologically active natural products.

EM-058 Ex-vivo Antioxidant Assay on Erythrocyte Model and Oxidative DNA Damage Protective Activity of Pithecellobium dulce Leaves Masuma Akter. Md. Mahadi Hasan, Md. Ekramul Islam*, Mst. Shahnaj Parvin Department of Pharmacy, University of Rajshahi, Rajshahi-6205, Bangladesh Correspondence: ekram.ru.ac.bd

As medicinal plants are suitable alternative for synthetic and chemical drugs hence this study was performed to evaluate the inhibitory effect of chloroform (CHF) and ethyl acetate (EAF) fraction from Pithecellobium dulce leaves extract on oxidative damage of DNA and human erythrocyte induced by AAPH [2,2’-azobis(2- methyl propionamidine) dihydrochloride]. It was determined that the oxidative DNA damage, hemolysis and lipid peroxidation of erythrocytes induced by the aqueous peroxyl radical of AAPH were suppressed by both CHF and EAF extracts of P.dulce in concentration- and time-dependent manners. Erythrocyte suspension at 5% hematocrit was incubated with PBS (control), or pre-incubated with CHF and EAF extracts of P.dulce at the concentrations (50-200µgl/ml) for 30 min. Then it was incubated with 50 mM AAPH at 2, 4 and 6 h at 37ºC. Where after 6 h EAF and CHF generate 7.35±0.18 and 8.80±0.72 pmol/g Hb MDA compared to AAPH (p<0.05) 24.51±0.23pmol/g Hb MDA. P.dulce extracts also prevented the decline of superoxide

179

dismutase (SOD) activity and the depletion of cytosolic glutathione (GSH) in erythrocytes. These results imply that P.dulce may have protective antioxidant properties.

EM-059 A survey on the treatment and medication used for gynecology and obstetrics department in a tertiary care hospital Simin Shabnam Lopa1*, Mst. Hajera Khatun1*, Farhana Hasan2, Saima Akter1, Md. Shahriar Nur1 1Department of Pharmacy, Varendra University, Rajshahi-6204, Bangladesh. 2Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh. 3Department of Pharmacy, University of Rajshahi, Rajshahi-6205, Bangladesh. *Corresponding e-mail: [email protected], [email protected]

Introduction: Gynecology is the branch of physiology and medicine which deals with the functions and diseases specific to women especially those affecting the reproductive system and obstetrics a branch of medical science that deals with pregnancy, childbirth and the postpartum period. The objective of this survey was designed to evaluate the gynecological complications, treatment procedure and patient’s awareness about drug use in our society. Methodology: It is a cross sectional study done in department of gynecology and obstetrics, Rajshahi medical college hospital, during the period of 1st February to 15 June , 2018. Data were collected purposively by Individual interview through a structured questionnaire. Results: Total 339 women of age group (15-45 years) were taken. Among them, 85.5% were within 18 to 40 years old, 43%were above 40 years old and 6% were bellow15 to 18 years old. Socio-economic status found low 37.8 %, medium 35.1% and rich 27.1% according to their family income. Mostly they were house wife (94.1%) and 51% of them are educated at primary level. About 49% women have no idea about contraceptive drug effect. 71.7 % of their husband was less educated; they drop the education at the school level. Among respondents most of the patient 63.3% came for delivery or pregnancy follow up, 13.9 % came with tumor and ovarian cyst. Rest of the patient came up with the complication of vaginal infection and miscarriage. The common physical complication found in pregnant women weakness, headache, nausea, vomiting, hair loss. 13.9 % patients had abnormal menstrual history. Majority women 49% were took hormonal birth control method. The antibiotics were prescribed frequently about 24%, analgesic drugs were

12%, 11% were H2 blocker, 7% were PPI (proton pump inhibitor) and vitamins. About 3% patients were prescribed for anti-cholinergic drugs. Among these antibiotics 39% were ceftriaxone sodium and 25% were metronidazole. Moreover 10% was found cephradine and flucloxacillin. The most of patients used ranitidine and a few patients used PPI. Conclusion: The reproductive health condition of women in our country is not at all satisfactory. Illiteracy still is in higher level. There is a lack of knowledge about contraceptive hormones those they are using frequently for birth control. A strong awareness counseling activity for the patient is in dire need in hospital level to improve the reproductive as well as women health condition.

EM-060 Testing homogeneity between raters Papia Sultana Professor, Department of Statistics, University of Rajshahi, Bangladesh

In biomedical sciences, education, psychology, sociology and also in other fields, two observers evaluate and classify subjects into two or more mutually exclusive categories, and researchers are interested to measure the agreement between the raters (or, observers) for such categorical or ordinal data. These raters can be persons, test methods, or any other proper tool depending on the problem. The most popular measure of agreement for this problem is Cohen’s Kappa. Although Cohen’s Kappa has meaning as percentage agreement corrected for chance, its application supports homogeneous raters only. Now, one challenge is to testing homogeneity of raters. In this paper, I have developed a probability model which allows us to test homogeneity between raters. Keywords: Agreement index, Cohen’s Kappa, Probability model, Homogeneity.

180

EM-061 Modeling on Monthly Maximum Chloride Content of Betna-Kholpatua River of Satkhira District in Bangladesh Shaima Jannat1*, Provash Kumar Karmokar1 and Dulal Chandra Nandi2 1Department of Statistics, University of Rajshahi, Rajshahi, Bangladesh 2Department of Statistics, Comilla University, Comilla, Bangladesh * Email of Communicating Author: [email protected]

Surface water salinity is a problem all over the country. About 30 percent of the cultivating land in coastal areas is affected by salinity at tidal flooding times. The rivers Betna and Kholpatua are flowing through Satkhira and Khulna areas are important in the south region of the country. These rivers continuously increase the surface water salinity by overflowing its bank that has been affecting the agricultural productivity of such areas. Although the Electivity Conductivity (EC) and Chloride content caused the water salinity both for surface and ground water, the Chloride content is more serious. Monthly Maximum Chloride in high tide (mmaxHT) has considered and has modeled using data of 33 years for the period of 1981 to 2013. An indication of having seasonality is noticed by time series plot of data which has confirmed by the formal HEGY test. As such five SARIMA models have been fitted and using the diagnostic tools and model selection criteria the best fitted model for the surface water salinity found as SARIMA(4,0,1) (0,1,1)6 for mmaxHT. Finally we have predicted the water salinity parameter up to the year 2025. It is observed from the predicted value that although the predicted values are increasing within the time period from January to May but for June it decreases. This could be a message for the water development board and concerning author as well as the Bangladesh government for their policy implication to reduce the danger of salinity in the southwest region of the country. Keywords: Surface Water, Chloride Content, Time Series Modeling, SARIMA.

EM-062 Climate Change Variables and its Impact on Crop Production of Barind Region of Bangladesh Mst. Afsana Ansari 1*and Provash Kumar Karmokar1 Department of statistics, University of Rajshahi, Bangladesh Email of Communicating Author: [email protected]

The adverse impact of climate change on the agricultural production is seen in different areas of the world. The climate change and its impact also affect the agricultural crop production of Bangladesh. Although the Barind region is called as store of crops, sometimes the production is hampered by negative impact of climate change. As such we are interested to know how crop production of Barind region is affected by the climate change variables for such region. Impact of climate change (temperature both in minimum and maximum and rainfall) and its effect on agricultural crop have examined by cointegration technique. As influential to the respective production the Engle-Granger cointegration methodology have applied to observe the long run relationships of production variables YpRanAu, YpRajAm, YpRanAm and YpRajWh with climate change variables (TxRanAuCC, TnRanAuCC and rRanAuCC), (TxRajAmCC, TnRajAmCC and rRajAmCC), (TxRanAmCC, TnRanAmCC and rRanAmCC) and (TxRajWhCC, TnRajWhCC and rRajWhCC) respectively. Cointegrating relationships (YpRanAu with the variables TxRanAuCC, TnRanAuCC and rRanAuCC), (YpRajAm with the variables TxRajAmCC, TnRajAmCC and rRajAmCC), YpRanAm with the variables TxRanAmCC, TnRanAmCC and rRanAmCC) and (YpRajWh with the variables TxRajWhCC, TnRajWhCC and rRajWhCC) are identified. For each cointegrating relations, ECM

model have computed and among them as a best representing model for YpRanAuCCt with climate change variables, we have recorded the absolute value of the speed of adjustment coefficient εˆ1,t−1 as 0.61776. This value is statistically significant at 5% level. Since this value is significant and it is less than one, the system is stable. The value of the speed of adjustment coefficient indicates that approximately 61.78 % of any deviation from the long–run path is corrected within production period. The Ljung-Box Q-statistics are statistically insignificant for the lag 4,8,12 and 16 indicating that there is no inadequacy of the ECM model. The R 2 is found to be 0.335 implies that the independent variables can explain about 33.50% of total variation. Keywords: Climate change, Barind region, Cointegration, ECM and Speed of adjustment coefficient.

181

EM-063 Prevalence, Awareness, Treatment and Control of Hypertension among Adults in Bangladesh: A statistical analysis to adjust cluster effect Mira Rani Debnath1, Papia Sultana2, Dulal Chandra Roy3 1M.Phil student, Department of Statistics, University of Rajshahi. E-mail: [email protected] 2Professor, Department of Statistics, University of Rajshahi. E-mail: [email protected] 2Professor, Department of Statistics, University of Rajshahi. E-mail: [email protected]

Frequently in public health research observations are nested within cluster. In such clustered data variation occurs not only at individual level but also at cluster level. Statistical models are available in literature for this type of clustered data to incorporate clustering effect in the model. But, researchers in this field are not passionate to such statistical models for clustered data rater to use simple models which are targeted to individual level analysis only. In this paper, we will describe statistical models to analyze cluster level variation along with individual level. We will demonstrate the necessity to use appropriate model using data from Bangladesh Demographic and Health Survey (BDHS) 2011 for measuring prevalence, awareness, treatment and control of hypertension among adults in Bangladesh. BDHS-2011 is the only survey that provides us biomedical parameters. It was a nationwide survey based on two-stage clustered sample. Hypertension is a growing public health problem in Southeast Asia, particularly in Bangladesh. Globally hypertension has been identified as a major health consequence which contributes to the risk of cardiovascular diseases (CVDs), stroke, kidney failure, disability, and premature death. It is a condition that occurs as a result of repeatedly elevated blood pressure exceeding 140 over 90 mmHg whereby a systolic pressure above 140 with a diastolic pressure above 90. However, normal blood pressure is below 120/80; readings between 120/80 and 139/89 is called pre-hypertension. Systolic blood pressure is the pressure in the arteries as the heart contracts and pumps blood forward into the arteries whereas diastolic represents pressure as a result to relation of the arteries after contraction. It has been called a silent killer as it is usually without symptoms. Diagnosis of high blood pressure is usually measured with a device called sphygmomanometer. Similar to other developing countries, obesity, diabetes, and hypertension are an emerging public health concern in low-income countries like Bangladesh, where a considerable proportion of people with hypertension are undiagnosed and/or untreated, due to limited access to healthcare and high treatment costs. According to the Global Burden of Disease 2010 study, hypertension has become the fifth leading risk factor for death and disability in Bangladesh. Some studies have reported the overall prevalence of hypertension in Bangladesh incorporating respondent-level analysis. Therefore, in this study we compared the results from respondent-level analysis only and results from cluster and respondent-level analysis. Keywords: Clustered data, multilevel model, hypertension.

EM-064 Multilevel analysis to predict both Tobacco smoking and smokeless tobacco product use in Bangladesh Moniruzzaman Monir1 Saroje Kumar Sarkar2 and Papia Sultana3 1Department of Statistics, University of Rajshahi, Bangladesh. E-mail: [email protected] 2Department of Statistics, University of Rajshahi, Bangladesh. E-mail:[email protected] 2Department of Statistics, University of Rajshahi, Bangladesh. E-mail:[email protected]

In tobacco use (both smoking and smokeless tobacco) has become not only a major contributor to the country’s high morbidity, but also the biggest drains to the national economy. The present research aimed to identify significant correlates of tobacco use (both smoking and smokeless tobacco) using multilevel logistic regression. We had used secondary data collected by the Global Adult Tobacco Survey (GATS), 2010. It had been observed that prevalence of both tobacco smoking and smokeless tobacco product use was 7.07%, of which 14.46% were male and 0.68% were female. From multilevel logistic regression, it had been found that respondents of older age group (above 55 years) respondents were 12 times more likely to use both tobacco smoking and smokeless tobacco product than youngest group (less than or equal to 24 years) (OR=12.14, 95% CI=7.73, 19.07). Respondents with no formal schooling were most likely to use both tobacco smoking and smokeless tobacco product among the respondents. Businessman and farmers were more likely to be both user than employee and retired/unemployed persons or worker/laborer were less likely to be both user that employee. Respondents of lowest household wealth index group were most likely to use both tobacco smoking and smokeless tobacco product among the respondents. Therefore, along with implementation of WHO Framework Convention on Tobacco Control, a nationwide campaign is needed to aware people in rural area about the health risks of both tobacco smoking and smokeless tobacco product use. Keywords: Tobacco smoking, correlates, multilevel logistic regression, GATS, Odds Ratio. Submission for Student Paper/Poster Competition? Yes

182

EM-065 Students’ Performance and its Technical Efficiency in the Secondary Education Level of Rajshahi, Bangladesh Md. Ayet Ali1* and Provash Kumar Karmokar2 1 National Academy for Education Management (NAEM), Dhaka-1205 2 Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh * Corresponding Authors Email: [email protected]

Quality education is a vital issue to implement the Sustainable Development Goals (SDGs) of Bangladesh. It is big challenge to ensure the quality education in different stages of the education system of the country. Quality education may focus on bettering students’ performances. A number of variables are responsible for the good performances of the students in the secondary level. Among them study time, class attendance, curricular and co-curricular activities and extra-curricular activities, family care, tuition are very important to reach the targeted output. Together with these professional skilled of teachers are important and that may be achieved through training programs. The trained teachers may execute such training outcomes in respective working places accordingly. In this study we are intended to measure the technical efficiency of the students’ performance on basis of the authors’ survey data of Rajshahi district of Bangladesh following two-stage random sampling methodology. The data of 625 students, 125 teachers and 40 head teachers are collected as respondents through questionnaires, observation schedules and observation check list. A written exam in English has taken only for the students of class nine by well structured question and the marks secured by the students are considered as the output variables in this study. According to the objectives we tried to identify the relevant factors/variables influencing both the technical efficiencies and inefficiencies. The highest technical efficiency is recorded as 72% while the average technical efficiency is found as 70% following the above mention error components variables and inefficiency variables. Finally teaching learning activities and its institutional short comings have been identified through the Stochastic Frontier Analysis (SFA) modeling for the study area. Keywords: Secondary Education, Students’ Performance, Technical efficiency, Sustainable Development Goal (SDG), Stochastic Frontier Analysis (SFA).

EM-066 A Comparison of Technical Efficiencies of Boro Rice Producers between Natore and Bogura Districts of Bangladesh Ranjan Kumar Kundu, Md. Rakib Pervej* and Provash Kumar Karmokar Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh. *Communicating Authors: [email protected]

Bangladesh is the fourth largest rice producing country in the world followed by China, India and Indonesia (FAO, 2015) and also the fourth biggest rice consuming country. Among the rice verities Boro is more contributory in the country and huge farmers are engaged with the production of such rice. As in the northern region Natore and Bogura districts are famous for Boro production our aim was to investigate the impact of input variables and identifies identifies factors that explain differentials in technical efficiency on the Boro producers. Finally, we are intended to make a comparison between the producers of selected districts. Cross sectional data were collected for the year of 2017-18 cropping year from 186 farmers, comprising 88 of Natore district and 98 from Bogura district of Boro rice farmers following multistage sampling technique. A production function using maximum likelihood method is estimated and efficiency scores of individual farmer are calculated using Stochastic Frontier Approach. The result found study indicates that, the Boro rice farmers of Natore district is more technically efficient with mean technical efficiency of 0.824 compared to 0.798 of farmers of Bogura district. The empirical evidence suggests that the elasticity of production to fertilizer cost, seed cost, irrigation cost, land size is positive and statistically significant for Natore district, while only seed cost and irrigation cost are positive and statistically significant for Bogura district, which therefore suggested that an increase in each of these variables could enhance Boro output. Further age, educational level, farming experience, mobile access, extension service and credit access have been identified as the contributory in efficiency model in both districts while only membership of co- operative is marked Natore district. The study recommended that Boro rice farmers of Natore district should increase the fertilizer cost, seed cost, irrigation cost, land size in order to improve production, at the same time as, Boro rice farmers of Bogra district should increase only seed cost and irrigation cost in order to improve production. Keywords: Technical Efficiency, Production Function, Stochastic Frontier Approach

183

EM-067 Patterning Prevalence and Correlates of Under-Five Child Malnutrition in Bangladesh Based on BDHS (2007-2014) Data Md. Tahidur Rahman1, 2, Noyon Ahmed3, Md. Jahangir Alam4, Dulal Chandra Roy5 and Papia Sultana6 1Assistant Professor Department of Statistics, Islamic University, Kashia-7003, Email: [email protected] 2Research Students, Department of Statistics, University of Rajshahi, Email: [email protected] 3Research Students, Department of Statistics, University of Rajshahi, Email: [email protected] 4Research Students, Department of Statistics, University of Rajshahi, Email: [email protected] 5Professor Department of Statistics, University of Rajshahi, Email: [email protected] 6Professor Department of Statistics, University of Rajshahi, Email: [email protected]

Malnutrition is referred to as the greatest single threat to the world‘s public health, especially for developing countries. Rates of malnutrition in Bangladesh are among the highest in the world, with six million children estimated to be chronically undernourished (DHS, 2011). The nutritional status of under-five children is the most sensitive indicator of a society, a country as well as the world’s public health status. This study investigated the differential impact of some demographic, socio-economic, environmental and health related factors on nutritional status. The objectives of this study are to find the prevalence and patterns of stunting, wasting and underweight among under-five children in Bangladesh from 2007 to 2014. Data and Methods: Secondary data of sample size 5300, 7647 and 6965 are extracted from Bangladesh Demographic and Health Survey (BDHS- 2007, 2011 & 2014) respectively. Indicator of malnutrition stunted (Height-for-age-z-score<-2), wasted (weight-for-height-z-score<-2) and underweight (weight-for-age-z- score<-2) were considered as the outcome variable. Along with descriptive analysis and factor analysis had been used to find patterns of malnutrition among under-five children in Bangladesh and its correlates. Results: Prevalence of stunted among the under-five children were 41.70%, 40.67% and 36.57%; that of wasted were 16.94%, 15.48% and 14.43%; and underweight were 39.79%, 35.80% and 32.45% in 2007, 2011 and 2014 respectively. From the factor analysis, it had been found that top most five correlates for stunting in 2014 were father’s education, household wealth index, mother’s education, age of child and duration of breast feeding, in 2011 those were household wealth index, mother’s education, father’s education, father’s occupation and frequency of antenatal visits during pregnancy; and in 2007 those were duration of breast feeding, age of child, father’s education, mother’s education and household wealth index respectively. Again top most five correlates for wasting in 2014 were father’s education, household wealth index, mother’s education, age of child and duration of breast feeding, in 2011 those were household wealth index, father’s education, mother’s education, father’s occupation and disparity in rural-urban residency; and in 2007 those were age of child, duration of breast feeding, age of child, father’s education, household wealth index and mother’s education, respectively. Further, those for underweight in 2014 were father’s education, household wealth index, mother’s education, age of child and duration of breast feeding, in 2011 those were father’s education, household wealth index, mother’s education, father’s occupation and disparity in rural-urban residency; and in 2007 those were age of child, duration of breast feeding, age of child, father’s education, household wealth index and mother’s education, respectively. Conclusion: This study revealed that father’s education level, wealth index, mother education, age of children and duration of breastfeeding are the most significant factors and these factors are direct influenced on the stunting, wasting and underweight of children. For improving the nutritional status of children government or policy maker should take necessary steps for controlling these factors by modifying some of the existed intervention strategy. Keywords: Child malnutrition, Stunted, Wasted, Underweight, Factor analysis.

EM-068 Generation of a new thermo-sensitive genic male sterile rice line by targeted mutagenesis of TMS5 gene through CRISPR/Cas9 system Hirendra Nath Barman 11,2, Zhonghua Sheng 22, Sajid Fiaz 32, Min Zhong 42, Yawen Wu 52, Yicong Cai 62, Wei Wang 72, Guiai Jiao 82, Shaoqing Tang 92, Xiangjin Wei 102* and Peisong Hu 112* 1Plant Physiology Division, Bangladesh Rice Research Institute, Gazipur-1701, Bangladesh. 2State Key Laboratory of Rice Biology, China National Rice Research Institute, Hangzhou 310006, China. Email: [email protected]

Developments of high yielding varieties are the core strategy to produce more and quality rice grain. Hybrid rice breeding with high yield potential is an alternative way to address demand of increasing population. Two-line hybrid rice with high yield potential is increasingly popular and the photo- and temperature- sensitive male sterile line is one of the basic components for two-line hybrid rice breeding. The development

184

of male sterile lines through conventional breeding is a lengthy and laborious process, whereas developing thermo-sensitive genic male sterile (TGMS) lines for two-line hybrid breeding by editing a temperature- sensitivity gene by CRISPR/Cas9 is efficient and convenient. Here, thermo-sensitive genic male sterility (TGMS) was induced by employing the CRISPR/Cas9 gene editing technology to modify the gene TMS5. Two TGMS mutants, tms5–1 and tms5–2, both lacking any residual T-DNA, were generated in the indica rice cultivar Zhongjiazao17 (cv. YK17) background. When grown at a sub-optimal temperature (22 °C), both mutants produced viable pollen and successfully produced grain through self-fertilization, but at temperatures 24 and 26 °C, their pollen was sterile and no grain was set. F1 hybrids derived from the crosses between YK17S (tms5–1) and three different restorer lines outperformed both parental lines with respect to grain yield and related traits. The YK17S generated by CRISPR/Cas9 system was proved to be a new TGMS line with superior yield potential and can be widely utilized in two-line hybrid breeding of indica rice.

EM-069 Transforming rice based cropping systems for achieving SDG 2.3 for Bangladesh rice M. Nasim, A. B. M. Mostafizur, M U Salam, T. L. Aditya and M S Kabir Bangladesh Rice Research Institute *Correspondence: [email protected]

Presently we have 17 sustainable development goals (SDG) on our hand to transform our future world to be more habitable and beautiful and to be achieved by 2030. The goal number 2 mostly works on agriculture. Important target of goal 2, directly related to agriculture production is 2.3, doubling the agriculture productivity and incomes. This target has 2 indicators, based on which the achievement of the target will be measured, 2.3.1. Doubling the volume of production per labour unit and 2.3.2. Doubling the average income of small-scale food producers. To achieve these difficult targets we need to transform our food production system, as crops are grown in cropping systems, we need to transform it. In this study we look into the future and draw a roadmap for Bangladesh rice by prediction` based on recent past trend of various data and information with assumption and justification to reach a decision on how transformation of cropping systems can bring about increase in rice productivity by 2030. Rice production in Bangladesh is steadily increasing after its independence in 1971 despite of having adversity of increasing population and limited resources like very low cultivable land/head, presence of problem soil and water scarcity and abundance etc. Green revolution: HYV, fertilizer and irrigation and associated technologies contributed to increase in rice production. Further increase is very challenging. In this regard pathways are constructed with major interventions include accelerating genetic gain of the future rice varieties, exploitation to achieve potential yield of a variety in real production niches, curtailing adoption lag of potential technologies, utilization of fallow period within cropping patterns by accommodation of an extra crop, expansion of rice cultivation in newly accredited Charlands, cost effective input use, mechanization etc. To achieve the targets of the goal drivers and draw backs were considered.

EM-070 Hermetic Bag: An Effective Storage Technology for Rice Seed/Hermetic Bag: A Key Element in Enhancement of Food Availability Mohammad Afzal Hossain*, MA Awal1, MR Ali1, and MM Alam1 *Corresponding author: PhD Fellow, email: [email protected] 1Professor, Dept. of FPM, Bangladesh Agricultural University

Hermetic bag (HB) has emerged as a significant alternative to other methods of storage. The study was conducted to compare traditional storage system along with HB bag in Bangladesh Agricultural University in completely randomized design with three replications of six treatments – Dole, Motka, Plastic Drum, Plastic Bag, HB i.e. Perdue Improved Crop Storage (PICS) and GrainPro Bag. The changes of moisture content (MC) of stored rice were found maximum in Dole. Because upper surface of it is open and paddy is hygroscopic. Whereas, MC remained same (12%) in GrainPro and PICS bags during five months. Since these are airtight and prevent moisture exchange with surrounding. Increased moisture, high temperature and relative humidity favored insect infestation in traditional storage technologies. Consequences, the highest storage loss of stored rice was found in Motka followed by Dole and plastic bag undoubtedly due to persistence of high MC and oxygen whereas no insect infestation was observed in HB. Due to porous behavior of Dole, Motka and Plastic bag permits serious loss by insects than that in HB. Germination capacity fell down to 42% in Dole. It was observed (≥ 92%) in HB. Highest financial benefit can be obtained using PICS Bag followed by GrainPro Bag and Plastic Drum. HB made insecticide use redundant in paddy

185

storage, maintain seed viability. However, HB may be adopted widely at farmers' level for ensuring quality seed and enhancing food availability. Keywords: Hermetic, Effective, Profitable, and Technology Submission for Student Paper? Yes

EM-071 Prediction of paddy quality stored in different technologies using artificial neural network model Mohammad Afzal Hossain*, MA Awal1, MR Ali1, Machbauddin2, and MM Alam1 *Corresponding author: PhD Fellow, email: [email protected] 1Professor, Dept. of FPM, 2Asst. Professor, Dept. of CSM, Bangladesh Agricultural University

Artificial neural network (ANN) offers machine learning and prediction in performing agriculture task smartly. It can identify subtle patterns in input training data which may be missed by conventional statistical analysis. A multilayer feed forward with back propagation of error learning mechanism was developed in MATLAB 2015 software to predict paddy quality stored in different technologies using ANN. The real time observations were used to train the ANN model obtained from storing paddy in different technologies. Collected data were separated into two sets i.e. training (75%) and test (25%). The ANN model was trained by backpropagation algorithm so that application of a set of input would produce the desired set of output. Three models of hermetically stored paddy were developed i.e. first one was for germination, second one was for price and third one for profit prediction of stored paddy. The developed models were 4-layered feed- forward neural network. Input layer of the models consist of 6-7 neurons, which corresponds to the input variables. The output layer has 1 neuron, which represents the predicted germination, price and profit, consecutively of the stored product. These showed a set of inputs, which self-adjust to produce consistent responses. Developed ANN models can predict germination (R2=0.674), price (R2=0.74) and profit (R2=1.00) of stored paddy quality as of similar finding of experimental data. Kew words: Prediction, paddy quality, stored, technologies, and model Submission for Student Paper? Yes

EM-072 Disease Resistance Breeding Program atBRRI MAI Khan1,*, MM Rashid1, M Khatun2, PS Biswas2, TH Ansari1, MA Latif1, Y Fukuta3 and MS Kabir4 1Plant Pathology Division, 2Plant Breeding Division,4Director General, Bangladesh Rice Research Institute, Gazipur- 1701, Bangladesh and 3Japan International Research Center for Agricultural Sciences (JIRCAS), Japan *Correspondence: [email protected]

Diseases are playing an important role for sustainable rice production in Bangladesh. Among the different diseases, blast, sheath blight (ShB), false smut (FSm), bacterial leaf blight (BLB) and tungro are now the most destructive and recurrent problems in changing climate.Use of host resistance is the best alternative for disease management but durable host resistance against biotic stresses is comparatively more difficult than abiotic stresses. Predictingthe present situation, BRRI initiated the disease resistance programme with the collaboration of national and international agencies like JIRCAS, IRRI, UPM etc.Regarding the blast and BLBresistance, BRRI has already studied the avr gene distribution among racesof these pathogen in Bangladesh and identified the candidate genes Pi9(t) for blast, Xa21 for BLB and tsv-1 for tungrousing differential system and molecular markers. These genes have also been introgressed into the popular rice varieties following marker assisted breeding (MAB) and pathogenicity tests for improving their genetic background in respect of different diseases.To mitigate the future challenges, BRRI has also initiated the research activities on the identification of novel QTLs or genes particularly for blast and tungro disease resistance from the local landraces. The high-throughput methods for the detection of single nucleotide polymorphisms (SNPs) are using in BRRI for high-resolution genetic map construction, linkage disequlibrium-based association mapping, genetic diagnostics, genetic diversity analysis, cultivar identification, phylogenetic analysis, and characterization of genetic resources. The modern biotechnological tool CRISPR/Cas9 is also practicing in BRRI for developing blast resistant variety. To speed up breeding cycle, now a days BRRI is practicing modified field RGA(rapid generation advance) and trying to harvest 4 generations in a year. As a result, some blast and BLB resistant lines are now in most advance stages and lots of lines are available in the breeding pipe lines. BRRI is very much optimistic to combat the disease problem by host resistance for sustainable rice production in Bangladesh.

186

EM-073 Edaphic Suitability Mapping of BRRI dhan50 and BRRI dhan72 Md. Abdullah Aziz, Md. Ismail Hossain,Rokib Ahmed,Niaz Md. FarhatRahman, Md. Abdul Qayum, Md. Abdullah Al Mamun, Md. Shahjahan kabir *Correspondence: [email protected]

Bangladesh is an agro-based developing country and rice is the main agricultural product. Most of the people consume rice three times every day. Agricultural land of Bangladesh dominantly covered by rice crop. Sufficient rice production is the key to ensure food security in Bangladesh. In fact, rice security is synonymous to food security in Bangladesh as in many other rice growing countries. In future more rice production will be required because of increasing population, on the other hand our land resource is decreasing.Thus we need to best use of limited land resource. Suitability maps to put land to the use for which it is best suite.Our land is not homogenous all over the country. Various physical and chemical properties of soil varies spatially, similarly various rice varieties are suitable for some specific physical and chemical properties. As we need to high production with limited land, so it will be very helpful if we have variety wise suitability map based on soil properties. BRRI dhan50 and BRRI dhan72 are very prospective rice varieties. BRRI dhan50 is an aromatic rice variety in Boro season and BRRI dhan72 is Zinc enrich varieties in T.Aman season. So, these two varieties suitability maps are very important. Geographic Information System i.e. GIS software is a powerful tool for handling spatial variability analysis and preparation of suitability map. Soil physical properties namely land type, top soil texture, soil permeability, soil salinity, and slope were considered to determine area suitable for growing respective rice varieties. Finally these parameters were weighted by the relative influence for suitability assessment and these were done by Arc GIS 10.1 software. BRRI dhan50 is suitable in south and eastern side of Bangladesh. BRRI dhan72 is suitable in western side of Bangladesh. Keywords: Suitability map, Soil properties, Geographic Information System.

EM-074 GGE biplot and Stability Analysis for Yield Performance of Rice Genotypes across Different Environments in Bangladesh Md Abdullah Al Mamun1*, Md Ismail Hossain1, Md Shahjahan Kabir3, Niaz Md Farhat Rahman4, Md Abdul Qayum4, Md Abdullah Aziz4 and Rokib Ahmed5 *Corresponding author’s email: [email protected] 1 Scientific Officer, Agricultural Statistics Division, Bangladesh Rice Research Institute (BRRI), Gazipur, Bangladesh 2 CSO & Head, Agricultural Statistics Division, Bangladesh Rice Research Institute (BRRI), Gazipur, Bangladesh 3 Director General, Bangladesh Rice Research Institute (BRRI), Gazipur, Bangladesh 4 Senior Scientific Officer, Agricultural Statistics Division, Bangladesh Rice Research Institute (BRRI), Gazipur, Bangladesh 5 Data Analyst, Agricultural Statistics Division, Bangladesh Rice Research Institute (BRRI), Gazipur, Bangladesh *Corresponding author’s email: [email protected]

Genotype-by-environment interaction (GEI) is an important phenomenon for analyzing adaptability and stability of rice genotypes under different climatic conditions. Thirty eight T. Aman rice genotypes were categorized as short, medium and long based on growth duration and arranged in a randomized complete block design (RCBD) with three replications. The present paper assess yield stability across different environments using the additive main effect and multiplicative interaction (AMMI) model and GGE Biplot analysis. Combined analysis of variance revealed significant genotypic, environmental effects and GXE interactions for yield and in total variation, environment contributed most of the part for different duration of rice genotypes. Locations Gazipur (E4), Satkhira (E8) and Sonagazi (E9) for long duration, Comilla (E3), Rangpur (E7) and Sonagazi (E9) for medium duration, Barishal (E1), Comilla (E3), Gazipur (E4) for short duration varieties were found more discriminating and informative among all. Ideal location were found Rajshahi (E6), Kushtia (E5) for medium duration varieties and Sonagazi (E9) for short duration. GGE Biplot indicated that the most promising genotypes based on high mean yield and stability were BRRI dhan30 (G7), BR10 (G3) for long duration, BRRI dhan32 (G26), BRRI dhan72 (G28), BRRI dhan49 (G24) for medium duration and BRRI dhan75 (G37), BRRI Hybrid dhan4 (G38), BRRI dhan66 (G35) for short duration varieties. Ideal genotype was found BRRI dhan30 (G7) for long duration, BRRI dhan32 (G22) for medium duration and BRRI dhan75 (G37) for short duration. So the identified genotypes in the present study could be used for commercial cultivation consistently and wider adaptation to particular environments for sustainable production. Keywords: AMMI, G × E interaction, GGE biplot, Rice Genotypes, Bangladesh

187

EM-075 Farm Machinery and Postharvest Technology Activities on Rice Production and Processing M. D. Huda, M. G. K. Bhuiyan, M. A. Hossen and M. A. Rahman Farm Machinery and Postharvest Technology Division Bangladesh Rice Research Institute, Joydebpur, Gazipur Email: [email protected]

Rice is the major cereal crop being grown widely and it is the main staple for the people of Bangladesh. In Bangladesh rice dominates over all other crop and covers 75% of the total cropped area and 92% farmers grow rice and have no more scope for horizontal expansion. So the vertical means is the only way to increase rice production. Vertical expansion includes the use of modern technologies such as high yielding varieties, use of quality seeds, seedling raising technique, adoption of plant protection measures, water management practices and farm mechanization from pre-harvest to post harvest operations. Since 1988, the government of Bangladesh has given emphasis to continue import of farm machinery and help mechanize farming activities. Increase cropping intensity, labour shortage at peak operation period is the reasons for increasing demand for appropriate farm machinery. Farm machinery and postharvest technology (FMPHT) division of BRRI developed more than 30 machines including rice processing technology. Some of machines and technologies are listed below.

EM-076 Region specific BRRI variety adoption: A simple way of increasing national production Md. Abdul Qayum1, Md. Ismail Hossain11 , Niaz Md. Farhat Rahman21 , Md. Abdullah Aziz31 , Md. Abdullah Al Mamun41, Rokib Ahmed51 and Md. Shahjahan Kabir61 1Agricultural Statistics Division, Bangladesh Rice Research Institute, Gazipur-1701, Bangladesh, *Corresponding author’s email: [email protected]

Abstract. ‘Food security’ is the synonym of ‘Rice security’. Increasing population, climate vulnerability and decreasing resources appeared as the great challenges to keep the pace of food production in Bangladesh. To overcome that challenges, we shown a technology in this study say ‘Region Specific Variety Adoption’ that can help to increase the national rice production in Bangladesh. To know the varietal performance at different season and region, replicated trails of RCB design with 40, 39 and 9 BRRI varieties set up in 10, 9 and 7 regions for Boro, T. Aman and T. Aus seasons respectively in Bangladesh. BRRI Recommended management practices followed to conduct the trials. From these trails we find out the potential yielded variety with yield, highest yielded variety with yield, Average yield and lowest yielded variety with yield. To compare the production performance of region specific BRRI variety, secondary data of area and production of rice from BBS were used in this study. We found, if we adopt the potential yielded variety at the respected season and region than total rice production will be increased 24.22%, in case of highest yielded variety it will be increased 11.03% but for wrong selection of variety (lowest yielded variety) production will be decreased more than 50%. Keywords: Food Security, Region, BRRI Variety, Adoption Submission for Student Paper/Poster Competition? No

EM-077 Weather Forecast based Rice Crop Management System in Bangladesh: An Experimental Approach for T. Aus Season Niaz Md. Farhat Rahman, Md. Abdullah Aziz, Md. Abdullah Al Mamun, Afroza Chowdhury, Rokib Ahmed, Md. Mofazzel Hossain, ABM Zahid Hossain, Mohammad Ashik Iqbal Khan, Md. Khairul Alam Bhuiyan, Md. Mozammel Haque, Tuhin Halder, Md. Ismail Hossain and Md. Shahjahan Kabir Agrometeorology and Crop Modeling Laboratory Bangladesh Rice Research Institute (BRRI), Gazipur-1701, Bangladesh (*Correspondence: [email protected])

Strengthening the early warning system to forecast extreme weather and provide action-oriented advisories is an essential step to increase rice yield as well as enhancing the farmer’s income. With this context, the study was conducted to assess the performance of weather forecast based rice crop management practices in comparison to traditional management. In this study, four districts during T. Aus season (April to August) were selected based on their diverse agro-climatic conditions: Rajshahi (cold & heat), Barishal (coastal flooding & waterlogging), Rangpur (cold and submergence) and Gazipur (favorable condition). Most popular and high yield potential varieties BR26, BRRI dhan48, and BRRI dhan65 were used in this regard.

188

Randomized Complete Block (RCB) design with three replications was followed as field experimental design. Also, grain yield data was adjusted to 14% moisture content. The weather research and forecasting (WRF) model was used to generate weekly weather forecasts on rainfall (mm), temperature (0C), humidity (%) and wind speed (m/s). Based on weather forecasting advisories were made and practiced into the respective experimental field of all locations through email communication. Linear mixed model (LMM), mean comparison test was performed for four locations to estimate the mean yield, mean square error, and coefficient of variation and find out the overall and location-specific performance of forecast-based management practices. Results revealed that throughout the T. Aus season weather conditions for all locations were more or less similar pattern. Rajshahi had the most extreme weather condition, always high temperatures and low rainfall and humidity. High rainfall was found in the Gazipur district. There is a good relationship between temperature and rainfall i.e. vice-versa. July is the highest rainfall and low-temperature month. On the other hand, May is the highest temperature and low rainfall month. LMM analysis showed there was a significant difference in management practices and the interaction of location and management practices meaning that yield of rice varieties varies in case of different management practices and yield also varies in different locations under different management practices. From the mean separation test, weather forecast based rice crop management practices performing better which produces 4.75 t/ha than traditional produce 3.91 t/ha. So, the research demonstrated high potential for field implications of the forecast based rice management practices resulted in 21.48% higher rice yield than the control plots with traditional management practices. Hence, analyses suggested weather forecast based rice crop management practices may be recommended for all locations in T. Aus rice cultivation having a similar environment as that of the experimental site. Keywords: Combined ANOVA, LMM, rice advisory, rice varieties, and weather forecast.

EM-078 Revision of the indicator 15.1.1 of the SDGs using satellite imageries Md. Abu Saleh1, Marzana Rahman Khuku2 1 Department of Statistics, University of Rajshahi, Kazla, Rajshahi-6206, Bangladesh. [email protected] 2 Institute of Environmental Science, University of Rajshahi, Kazla, Rajshahi-6206, Bangladesh. [email protected]

Spectral response of the objects upon the earth surface is the principal data collected continuously by the earth-orbiting satellites. This study had been carried out with the help of Landsat 8 imagery data for the interpretation of the Normalized Difference Vegetation Index (NDVI) imagery to find out how much forestry is present in any country or its particular region. NDVI is calculated from the visible and near- infrared light reflected by vegetation. The NDVI values range from -1 to +1 which is classified in 03categories as lower values of NDVI (0.1 and below) relate to barren areas of rock, sand, or snow, Modest values signify shrub and grassland (0.2 to 0.3) and high values designate moderate and tropical rainforests (0.6 to 0.8). The indicator 15.1.1 may be revised as accepting the mean NDVI value for a certain the region must lie from 0.58 to 0.84 whereas the current indicator is 20% forest area for the plain region and 33% for the geographical regions of the total area of any country according to this study. This revision is necessary because this will promote the research interest in satellite imageries among them who never used this technology before rather had used manual study and data collection which needed more fund and time consequently. Keywords: SDG 15, NDVI, Landsat 8. Submission for Student Paper/Poster Competition? Yes

EM-079 The Need Assessment of English Language in Tourism and Hotel Management in Bangladesh: A Statistical Approach Md. Shaon Akter, Dr. Abdullah Al Mamun, Dr. Muhammed Asgor Hussain, Dr. Md. Ayub Ali Lecturer, Department of English, Khwaja Yunus Ali University, Sirajgonj. Associate Professor, Department of English, Jashore University of Science and Tecnology. Associate Professor, Department of English, Islamic University, Kushtia. Professor, Department of Statistics, Rajshahi University. *Corresponding Email:[email protected]

Background and objectives: Tourism and hotel management is now the most felt and discussed sector in the world as well as in Bangladesh. Tourism and hotel management is such a business which is mostly depended on language, but there is no such report that demonstrates the need assessment of English in tourism and hotel management in Bangladesh. Thus, we aimed to identify the need assessment of English

189

language in tourism and hotel management in Bangladesh for the first time. Methods: We employed a statistical approach and a set of questionnaires which was used for collecting data. We asked them all the questions accordingly and put data on the paper. We collected the data from the owner and employees of the hotels and restaurants. Results: we have used statistical methods to analyze the datasets which reveals the poor proficiency in listening (0.25±0.58), speaking (0.27±0.64), reading (0.25±0.588), and writing (0.275±0.64) skills the problems of English language of the owners and employees are identified. Functional analysis of English language shows that the hotel employees at their workplace can give general information (62.50%) followed by providing services, offering help, asking for cooperation and answering question & solving problems are same (17.50%). Conclusion: In the present study, we performed an in-depth statistical approach to assess the needs of English language for hotel and tourism management industries in the context of Bangladesh. The different level of analysis showed that most of the employees contain poor level of communication skills in English that is why they are unable to conduct with the foreigners and to run the tourism and hotel business. In this way, we identified the problems and needs of English language among the employees in tourism and hotel management to improve the prospects of hotel and restaurant business in Bangladesh. Keywords: English language, tourism, hospitality management, needs assessment.

EM-080 Predictors and Risk Factors of Eating Disorder among Students in Malaysia Wah Yun Low1, Wen Ting Tong2, Md. Ashraful Islam3, Yun Li Chan2, Wan Yuen Choo Claire4 and Abdullah Adina2 1Dean’s office, Faculty of Medicine, University of Malaya ([email protected]) 2Department of Primary Care Medicine, Faculty of Medicine, University of Malaya 3Research Management Centre, Faculty of Medicine, University of Malaya 4Department of Social and Preventive Medicine, Faculty of Medicine, University of Malaya, Kuala Lumpur, Malaysia

This paper aims to investigate the predictors and risk factors of eating disorder among university students in Malaysia. A stratified cluster sampling was used to recruit participants from twelve faculties, two academies and two centres in a premier university based in Kuala Lumpur. This cross-sectional study was conducted among 1017/1132 students (response rate: 89.8%). The self-administered questionnaire was a combination of the Eating Attitude test-26 (EAT-26) and items related to perceived body weight, BMI, trying to weight loss, tobacco use, PTSD and depression (CES-D-10). The results showed that 13.9% of the university students were at risk of eating disorder. Prevalence of depression was 29.4%, PTSD 10.8%. Some 17.6% perceived their body weight to be underweight and 11.9% perceived themselves as overweight. Almost half (47%) of the students were trying to lose weight. Among those having eating disorder, a higher percentage was found among the female (14.8%), of Malay ethnicity (51.8), well off (61.0%), reside in the campus (81.6%) and perceived themselves to be of normal weight (47.5%). Results of the logistic regression showed that students who tried to lose weight predicts eating disorder. Eating disorder among university students merits attention. Efforts to promote healthy lifestyle are needed particularly focusing on students who are trying to lose weight. Keywords: eating disorder, University students, prevalence, risk factors

EM-081 Identification of hub genes and pathways in endometrial cancer via a bioinformatics approach Md Foyzur Rahman1, Md Rezanur Rahman1,*, Tania Islam2, Toyfiquz Zaman1, Md. Tofazzal Hossain3, Md Rezaul Karim1, Md Ayub Ali4 1Department of Biochemistry and Biotechnology, School of Biomedical Science, Khwaja Yunus Ali University, Enayetpur, Sirajgonj, Bangladesh 2Department of Biotechnology and Genetic Engineering, Islamic University, Kushtia, Bangladesh 3Department of Biochemistry and Molecular Biology, University of Rajshahi, Rajshahi, Bangladesh 5Department of Statistics, Faculty of Science, University of Rajshahi, Rajshahi, Bangladesh, Bangladesh *Correspondence E-mail: [email protected] or [email protected] (M.R.R.)

Endometrial cancer (EC) results from the abnormal growth of the cell lining of the uterus, and according to the recent World Health Organization (WHO) report, it is one of the most common gynecological malignancies of post-menopausal women worldwide. However, mechanisms that underlie to EC are still poorly known. Therefore, we aimed to identify potential biomarkers and networks among them of EC to clarify the potential mechanisms triggering it. We downloaded three mRNA datasets of EC and normal tissue samples (GSE17025, GSE115810, and GSE36389) from the Gene Expression Omnibus (GEO) to identify highly expressed differentially expressed genes (DEGs) in EC tumor tissues. Then, we validated

190

these DEGs using data from The Cancer Genome Atlas (TCGA) and subjected to gene ontology (GO) and pathway analysis to interpret their roles in biological processes and molecular pathways. We used STRING to construct protein–protein interaction (PPI) network to identify key hub proteins. We also performed survival analyses on the hub proteins to assess the prognostic values using “Survexpress”. A total of 62 aberrantly expressed genes (DEGs), of which 50 upregulated and 12 down regulated, were identified in EC samples from both the GEO and they were involved in various intricate molecular pathways e.g., cell cycle and cancer-associated pathways and also enriched in significant biological processes such as chromosome segregation, protein binding, cell division and proliferation etc. We detected six key hub proteins (EZR, NR4A1, SOCS2, VAV3, TCF4, MYO6) based on topological analysis of PPI network which play vital roles in the progression and regulation of EC. In conclusion, we have identified molecular biomarkers and pathways in EC which may provide new visions of understanding the underlying molecular mechanisms of EC that might help future studies. Keywords: Endometrial cancer, transcription factor, bioinformatics analyses, biomarkers, gene ontology, candidate drugs.

EM-082 Biological and biochemical characterization of Daboia russelii venom from Bangladesh and its neutralization efficacy by polyvalent antivenom Masum Miah1, Sohanur Rahman Sohan1, Mahedi Hasan1, Kazi Md. Faisal Hoque1 and Md Abu Reza1* Molecular Biology and Protein Science Laboratory, Department of Genetic Engineering & Biotechnology, University of Rajshahi, Rajshahi-6205, Bangladesh. *Correspondence: Professor Md Abu Reza, PhD, E. mail: [email protected]

Russell’s viper (Daboia russelii) is one of the most dangerous snakes in Bangladesh. Some recent bit incidence and causalities by Russell’s viper posed a serious concern to the health professionals. Recently it has been found in nine districts (Dinajpur, Chapai Nawabganj, Rajshahi, Naogaon, Natore, Pabna, Rajbari, Chuadanga and Patuakhali), although a few years back it was presumed that it has been extinct from Bangladesh. Here, we have carried out an in depth study on Russell’s viper venom. Snake venom was collected from local area in Rajshahi division with the help of local expert snake handler of Snake Venom and Rescue Center (SRCC) of Darusha, Rajshahi. LD50 of the venom was determined intraperitoneally (I.P.) in mice which was 0.3 mg/kg. However, the reported LD50 of Russell’s viper venom is 0.40 mg/kg I.P. Our LD50 value indicates that the Rullell’s viper venom from Bangladesh is more potent and may have different composition of venomous proteins. For biochemical characterization we have carried out the SDS page profiling, hemolytic activity, hemorrhagic activity, PLA2 activity and edema inducing activity. We also have determined the effect of the venom in muscle, skin, liver and kidney tissues of mice. Significant histolopathological alterations were observed caused the venom of Russell’s viper venom. We also have performed the notarization assay of Russell’s viper venom by Vins polyvalent antivenom. Interestingly, we found that Vins antivenom has low efficacy for neutralization of Russell’s viper venom. Our study suggests the necessity of developing of new antivenom for proper snakebite treatment.

EM-083 Factors Affecting Nutritional Status among the Kinnaura in Himalayan Region through Statistical Data Analysis Dr. Swapan Kumar Kolay Associate Professor & Head, School of Anthropology & Tribal Studies, Bastar University Dharampur Jagdalpur, District: Bastar 494001 Email: [email protected]

This study assesses the nutritional status of Kinnaura preschool children, identified the influencing factors and estimated the degree of nutritional status index. Conducted in nine villages selected from three different tehsils of Kinnaur district, Himachal Pradesh, the study used data from selected rural households and anthropometric measurements of preschool children. Household data were collected using structured questionnaire. The MS Excel software was used to calculate nutritional status indexes in which height-for- age was conducted to assess the influence of the explanatory variables on nutritional status. Results revealed that the proportions of children with either moderate or severe nutritional problems, about 45 percent boys and 39 percent girls suffer from mild to moderate forms of malnutrition according to weight for height where as 43 percent of boys and 47 percent of girls suffer from mild to moderate forms of under nutrition with

191

respect to height for age. Similarly, 62 percent of girls 68 percent of boys suffer from mild to moderate forms of malnutrition with respect to weight for age. In the present study boys are found to be more vulnerable as compare to girls in weight for age. Dietary consumption, mother’s education, mother’s position among housewives and child’s height were positively related to the child’s nutritional status. Also, mother’s age, child’s age and dependency ratio had negative influence on nutritional status. Policy options that would promote formal education for women, home use of nutritional diet and reduction in dependency ratio are recommended. Keywords: Nutritional status, Malnutrition, Kinnaura, Kinnaur district, Himachal Pradesh.

EM-084 Effects of road traffic accidents on travel in Puntland state of Somalia Prof. Dr M Sayedur Rahman and Abdulkadir Mohamoud Shire Department of Statistics, University of Rajshahi, Bangladesh

The rising of road accident cars is an increasing concern in Puntland travel roads, road loss is a major cause of both life and property. In recent years the road accidents become the popular problem which thousands of citizens and many wealthy were lost. The purpose of this study is to investigate the transportation system and proximity accidents. Using time series analysis, this study analyzed the accidents and transportations difficulties in Puntland from 2014 up to 2018. The proximity of accidents to Puntland transportation system, lack of skills and experience of drivers also the road is over and being broken all these were found to play the greatest role in facilitating transportation problems. This study definitely answers the question regarding correlation between proximity of Puntland travelling roads and poor skilled drivers to accidents, this is done using R programing software. The findings have implications for and treatment, road safety, training the drives and building resilient roads for travelling.

EM-085 Factors associated with malnourished among children under five in Bangladesh using generalized Poisson regression model Md.Saimul Islam1, KamrunNahar Koly 23*,

1 Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh 2 Department of Public Health, North South University, Dhaka-1229, Bangladesh 3Centre for Global Mental Health, Institute of Psychiatry, Psychology and Neuroscience, King’s College London- SE5 8AF, UK

Background: This study attempted to identify factors associated with the families having more than one any form of undernutrion children age under 5 years old. Methods:This paper used data from Bangladesh Demographic and Health Survey(BDHS-2011), which is based on a two-stage stratified sample of households. The data was analysedwith 6254 response of child age under five in a household and the cases was the children with moderate to severe malnutrition with z- scores < −2SD from the median of WHO reference. Results: Among study participants, 73.7%, 72.1% and 72.11% of household had at least one stunted, wasted and underweight child respectively. Result showed that the children living in the rural areas (IRR=1.077, CI=1.051,1.105, p<0.001), mother primary education (RR=1.041,CI=1.012,1.071, p<0.001), father had no formal education, RR=1.044, 95% CI=1.010,1.078) associated with stunting. Determinants of wasting were poorer SES,RR=1.078, 95% CI=1.024,1.135),father occupation (agro based worker, RR=1.068, 95% CI=1.006,1.135),toilet facility (Non improve ,RR=1.032, 95% CI=1.001-1.065),mother age 20-30 years. RR=1.033,95% CI=1.001,1.066). As besides, rural area (RR=1.077, 95% CI=1.044-1.112), mother education ( illiterate ,RR= 1.052,95% CI=1.004-1.102),mother age <21 years. (RR=1.272,95% CI=1.204- 1.344) associated with the underweight status of the children. Conclusions: The findings of this study have also confirmed the significant association of acute malnutrition with parents low educational status, mother poor nutritional status, poorer economic conditions in a family, number of ever born children, lower age of mother. Besides implementation of programmes focusing on mother nutritional knowledge, community based nutrition programmes should be enhanced to alleviate children’s malnourished conditions. Keywords: Malnutrition; Under-five children; Predictors; Generalized Poisson regression model; Bangladesh

192

EM-086 Impact of maternal anemia and household wealth index on children anemia in Bangladesh: Multilevel logistic regression analysis Md. Kamruzzaman1, Abu Yusuf, Abu Sayed Md. Al Mamun1, and Md Golam Hossain1 1Health Research Group, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh

Anemia is a public health problem especially among children in developing countries and it is one of the important predictors for growth and development. The aim of this study was to determine the impact of maternal anemia and household wealth index on anemia among pre-school children (6-59 months) in Bangladesh after removing cluster effect and after controlling the effects of selected confounders on children anemia with 95% confidence interval (CI). Data for this study was extracted from the 2011 Bangladesh Demographic and Health Survey (BDHS-2011). BDHS-2011 used a two-stage stratified cluster sampling technique for selecting households. The anemia status of pre-school children was the outcome variable of this study. Chi-square (χ2) test and two level logistic regression model were used in this study for finding the impact of maternal anemia and household wealth index on children anemia. Population attributable fraction (PAF) was utilized for calculation the contribution of maternal anemia and household wealth index on children anemia. Parents’ socio-economic, demographic, children age, birth weight, nutritional status were considered as cofounding variables. The total number of 2071 children aged 6-59 months were considered to examine their anemia status and it was observed that the prevalence of child anemia being overall 52.10%, of which 48.27% where in urban environments and 53.73% in rural areas in Bangladesh. The prevalence of mild, moderate and severe anemia was 29.60%, 21.70% and 0.80% respectively among anemic children in this country. The two level logistic regression model demonstrated that anemic mothers’ children were more prone for getting anemia than non-anemic mothers’ children (p<0.01). When controlled the effect of other variables, it was also found that anemic mothers were more likely to get anemic children (p<0.01). It was noted that children living in poor family who was more probable to get anemia than who lived in rich family (p<0.01). After controlling the effect of other variables, it was found same result but comparatively lower odds ratio [Adjusted odds ratio (AOR)=1.708]. When we compared middle family children with rich family, it was found that children who lived in middle family had more chance to have anemic than rich family children for both unadjusted and adjusted (p<0.01). Finally, population attributable risk percent (PAR%) demonstrated that the individual contribution of anemic mothers for getting anemic children by 20.65% for unadjusted, and for adjusted it was by 19.35%. In case of unadjusted, the individual contribution for poor family among their children anemic was 18.20% and for adjusted was 17.70%. It was found the contribution of middle family for having anemic children by 4.65% and 4.43% for unadjusted and adjusted cases respectively. A high prevalence of anemia among pre-school children (age, 6-59 months) was found in Bangladesh. This study suggests the presence of modifiable factors associated with child anemia. Our selected multilevel model provided that anemic mothers and poor family children were vulnerable for getting anemia in Bangladesh. These two groups are correlated, and PAF shows that these two groups together can put contribution by more than one third (38.85%) for children anemia in this country. Consequently, anemia can be considered as the major problem for children in Bangladesh, and Government and Non-government health organizations can consider our findings for societal intervention to reduce its incidence. Keywords: Anemia, Pre-school children, Bangladesh, Prevalence, Multilevel regression

EM-087 Quality Education Data for Achieving SDG4 Targets in Bangladesh: Prospects and Challenges. Dr. AQM Shafiul Azam, Chief DLP Division, BANBEIS, Ministry of Education 1, Zahir Raihan Road (Palashi-Nilkhet), Dhaka, Bangladesh

Bangladesh has taken substantive steps towards implementation of Sustainable Development Goal 4 (SDG4). Follow-up and review based on robust monitoring, reporting and evaluation of policies, systems and tools are essential for the achievement of SDG4. Thus, countries are required to address the critical challenges affecting the country's ability to monitor the progress of targets under the goal and to consolidate the national education statistical system and strengthen national education data. Given the challenges of SDG4 implementation, Bangladesh Bureau of Educational Information and Statistics (BANBEIS) is developing National Strategy for the development of Education Statistics and Action Plan (NSDES & AP) to establish this consolidation with a medium-term vision and also to develop national capacities to generate quality education data for monitoring the SDG4. The development of NSDES is for monitoring, evaluation and reporting in education to inform progress towards the achievement of Education 2030 both at national and international level and AP to be developed for improving data quality

193

and developing national capacities for generating and reporting education data. The development of NSDES is based on National Indicator Framework (NIF), Data Mapping and Action Plan (DM & AP) and Data Quality Assessment Framework (DQAF) prepared by ensuring participation of all relevant stakeholders under the leadership of two ministries of education in Bangladesh, Ministry of Education and Ministry of Primary and Mass Education. The NIF is the complete set of monitoring tool developed to fulfill the statistical demand for tracking the progress of implementation of SDG4 as well as our national mandates.

EM-088 Minimax Estimation of the Scale Parameter of the Laplace Double Exponential Distribution Mira Rani Debnath1, Md. Ayub Ali2, Dulal Chandra Roy3 and Papia Sultana4 1Research Fellow, Department of Statistics, University of Rajshahi, 2Research Fellow, Department of Statistics, University of Rajshahi, 3Professor, Department of Statistics, University of Rajshahi, Rajshahi-6205, BD 4Professor, Department of Statistics, University of Rajshahi, Rajshahi-6205, BD Mail address of the corresponding author: [email protected]

In this paper, minimax estimators of the scale parameter of the Laplace double exponential distribution have been developed under modified linear exponential (MLINEX) loss function applying the theorem of Lehmann(1950). The efficiency of the estimators are also been studied. Key words: Minimax estimator, Laplace double exponential distribution, Modified linear exponential loss function.

EM-089 Innovative Data Approaches for the SDGs: Bangladesh Perspective Md. Alamgir Hossen Deputy Director and Focal Point Officer, SDG Cell, Bangladesh Bureau of Statistics Email: [email protected]

Timely and reliable official statistics are the lifeblood of policymaking as they comprise an integral part of designing, monitoring, and evaluating development priorities. The data for monitoring SDGs need to be of high quality, accessible, timely, reliable, and disaggregated by income, sex, age, race, ethnicity, migration status, disability, geographic location, and other characteristics relevant in national contexts. The requirements of disaggregated information on population groups extend the main focus on the far side national trends and averages towards distinguishing subgroups of the population that were unnoticed with relevance specific development targets. Disaggregated statistics add values as a vital inputs for an additional concentrated policy-targeting. On the other hand, non-availability of timely and high-quality disaggregated data will, however, render the monitoring of progress towards the SDGs ineffective. Bangladesh Bureau of Statistics (BBS), the National Statistics Office (NSO) of Bangladesh, plays a lead role in the generation and dissemination of official statistics which can be treated as essential public good, typically through traditional methods of statistical surveys, censuses, and compiling data from administrative sources. While both national and global data needs are continuously expanding, the proliferation of new types of data is on the rise in the form of satellite and other digital images, digital records, machine-generated data, social media data, internet-compiled data, and consumer databases that provide an unprecedented opportunity for a more holistic, inclusive, and highly energized era of networked problem solving. These fast-paced changes in information and communication technology have led to the calls for a “data revolution” wherein the richness of conventional data sources and analytical techniques commonly used by the National Statistics Offices to produce official statistics. It needs to be complemented by new and innovative forms of data, and application of state-of-the-art statistical methods to produce more frequent, disaggregated, and high-quality data demanded by the users. The technology environment has continued to advance in recent years in Bangladesh as well to a stage that it appears now promising to harness big data for both, the achievement of SDG targets as well as the estimation for SDG indicators. There are some challenges which should be taken care also. It includes the size and volume of new types of data such as satellite images and mobile phones data are significantly bigger than conventional census and survey data sources that BBS is used to working with. Furthermore, the representativeness of new data sources needs to be carefully analysed as they may be focused to specific segments of the population only, unlike conventional data sources that are usually representative of the general population. To capitalize on the rich information provided by these non- conventional data sources, National Statistical System of Bangladesh will require critical and specific technical assistance in developing or enhancing the technical skills and infrastructure suitable for compiling,

194

validating, synthesizing, analysing, and disseminating data collected through traditional surveys, censuses, and administrative sources and complement these with non-conventional forms of data. In Bangladesh, data can now be collected easily using smart mobile phones to provide location-specific information on government facilities, water points, and environmental challenges. Bangladesh Bureau of Statistics (BBS) has already adopted the modern technologies in their data collection and processing process. Exploring the chance of sourcing from big data technology to deal with the large increase within the amount of knowledge that must be generated. BBS has started working to identify the areas where capacity building for Big Data analysis is required. As such, BBS is keen to develop and strengthen partnerships with different implementing agencies and all relevant stakeholders in the National Statistical System, including custodian agencies of SDGs indicators and other relevant international communities. To adopt the new technologies in generating official statistics, BBS is looking forward to technical support from them. As part of such initiatives, BBS has already initiated process for data mining and coordinating all the data producing agencies and development partners for monitoring SDGs. Statistics Act, 2013 mandated BBS to generate official statistics and validate, authenticate and store other official statistics generated by different national data providers. SID is providing policy guidance and support to BBS to expedite such efforts focusing the integration of big data into official statistics and mainstreaming the administrative data with official statistics to comply with the fundamental principle of official statistics adopted by United Nations.

EM-090 How Do Education and Skill Development Affect the Unemployment, the Earnings, and the Quality of Employment Status in Bangladesh Muhammad Ariful Islam Deputy Director, Bangladesh Bureau of Statistics

This paper examines the impact of education and skill development on the three types of labor market outcomes in Bangladesh: unemployment, earnings, and job quality. This paper investigates how quality education (SDG4) affect the quality employment (SDG8) in Bangladesh. This paper uses the Quarterly Labor Force Survey (QLFS) 2016-17 data of Bangladesh conducted by the Bangladesh Bureau of Statistics (BBS). Using the Linear Probit model (LPM), Probit model, Logit model and the Two-Stage Least Square (TSLS) method, this paper shows that education affects the unemployment status in a non-linear fashion. The study finds that education and skill development both increases the monthly earnings in Bangladesh. The instrument variable of parent’s schooling has strengthened the outcomes for education which implies that parent’s education does not favorably bias the outcomes. Using the multinomial logit model, this study also shows that education and skill development increases the quality or the standard of job in Bangladesh.

EM-091 Machine learning and bioinformatics models to identify gene expression patterns of glioblastoma associated with disease progression and mortality Zakia Zinat Choudhury1*, Utpala Nanda Chowdhury1, M. Babul Islam2, Mohammad Ali Moni3 1Department of Computer Science and Engineering, University of Rajshahi, Bangladesh. 2 Department of Electrical and Electronic Engineering, University of Rajshahi, Bangladesh. 3School of Medical Sciences, Faculty of Medicine and Health, The University of Sydney, Australia. *[email protected]

Glioblastoma (GB), the most frequent and aggressive form of malignant brain tumor, causes rapid death. The genetic mutation is the main reason for this fatal disease. Henceforth, it is essential to identify the causal genetic targets associated with GB survival. Plenty of publicly accessible gene expression and clinical data for GB patients from the Gene expression omnibus and the Broad Institute Cancer Genome Atlas (TCGA) dataset can allow us to study patient fatality prediction and thus to identify new GB biomarkers. In this study we applied bioinformatics and network-based approach to identify the altered genes associated with the GB comparing with normal mRNA expression data from the brain tissues. Total of 325 genes were found as differentially expressed in GB. Gene set enrichment analysis through protein-protein interaction (PPI), gene ontology (GO) and KEGG pathways also revealed their significance. We then applied machine-learning approach incorporating Cox Proportional Hazard models to the both clinical and RNA-Seq datasets to determine target biomarkers that affect the survival of the GB patients. Three genes (PHLDA1, IQGAP2 and SPARC) were identified by using univariate approach that have a significant effect on the GB progression and mortality. Thus, our combined Machine Learning and Bioinformatics approach revealed the target signature genes for GB progression that could be useful to develop potential drug targets for the GB.

195

EM-092 Prevalence of Hypertension and its Associated Factors among the Patients Admitted in Medicine Ward of Islami Bank Medical College Hospital, Rajshahi 1 2 3 4 Mst. Sharmin Sultana, Dr. Md. Anayet Ullah, Dr. M. Firoz Alam and Md. Monimul Huq 1 MPH Student, Institute of Biological Sciences, University of Rajshahi, Rajshahi, Bangladesh. 2 Professor and Vice Principal, Dept. of Community Medicine, Barind Medical College, Rajshahi, Bangladesh. 3 Professor and Director, Institute of Biological Sciences, University of Rajshahi, Rajshahi, Bangladesh. 4 Professor, Department of Statistics, University of Rajshahi, Rajshahi, Bangladesh. Correspondence: [email protected]

Hypertension (HTN) is a basic public health challenge globally, specifically in developing countries, and is defined as a persistent raised systolic or diastolic blood pressure equal to or more than 140/90 mmHg in adults aged 18 years and over. The study age group selected has been adults aged 18-69 years. Hypertension leads to cardiovascular diseases, stroke and kidney failure and is the leading cause of mortality and morbidity globally. The risk factors for hypertension, which can mostly be prevented through simple health promotion and preventive measure, are mostly known. The main objective of this study was to assess the prevalence and factors associated with hypertension among the patients admitted in medicine ward of Rajshahi islami bank medical college hospital. A cross-sectional study was conducted and simple random sampling technique was used for socio-demographic variables and lifestyle-related data collection using a pre-tested structured questionnaire. HTN (using WHO guidelines) was defined as systolic Blood pressure (SBP) ≥ 140 mmHg and / or Diastolic Blood pressure (DBP) ≥90 mmHg. The prevalence of HTN and controlled HTN were 37% and 16.33% respectively, among the sample of 300 patients. About 19% patients were aware of their condition. The age, gender, tobacco consumption, physical activity, habit of eating egg and family history of hypertension has significantly influenced to the general rate of hypertension (p<0.0001) using chi-square test. The patients of older age (60-65year) (p = 0.00) and a positive family history of hypertension (p=0.00) were associated with hypertension. Tobacco consumption [AOR = 2.04; 95% CI (1.47, 2. 38)], habit of eating eggs [AOR = 1.13; 95% CI (0.76, 1.69)], and any thyroid disease [AOR = 1.74; 95% CI (0.70, 4.34)] were factors associated independently with hypertension. The prevalence of hypertension moderately high among the patient who admitted in medicine ward at islami bank medical college hospital, Rajshahi and BMI was the strongest modifiable risk factor. Prevalence of hypertension among the patients was high. There is an urgent need for implementation of a comprehensive cardio vascular disease prevention program. Routine blood pressure measurements should be taken to improve the detection, prevention and treatment of hypertension. Hypertension prevention should focus on awareness generation, health promotion based approach. Required action should be taken to prevent hypertension by the government, non-government organizations (NGOs) and civil societies.

EM-093 Factors Affecting Neonatal and Infant Mortality in Bangladesh Susama Tanzim1, Farzana Yesmin1, Most. Farjana Haque1, and Md. Monimul Huq2 1 4th year honours student, Department of Statistics, University of Rajshahi, Rajshahi, Bangladesh. 2 Professor, Department of Statistics, University of Rajshahi, Rajshahi, Bangladesh. Correspondence: [email protected]

Reducing neonatal and infant mortality rate is challenge in Bangladesh. Although the neonatal and infant mortality rate has been reduced over time in Bangladesh keeping the country on track to reach SDG, the rate is still very high. The objective of this study was to identify factors affecting with neonatal and infant mortality in Bangladesh which would be an advanced step in effective policies to tackle the issue. The data used for this study was extracted from Bangladesh Demographic and Health Surveys (BDHS) 2014 based on a two-stage stratified sampling of households. Bivariate analysis was used to examine the differentials in neonatal and infant mortality. Multilevel logistic regression was employed to determine the factors affecting with neonatal and infant mortality. In this study we observed that for individual level, education for mother and father, breastfeeding, delivery factors, postnatal care factor played a dynamic role in neonatal and infant mortality. On the other hand place of residence, presence of qualified doctor and availability of family planning methods were found to be significant in community level. Public health programs need to adopt a comprehensive strategy to address the individual, and community level factors associated with neonatal mortality in Bangladesh. The study also highlights the importance of strengthening proper postnatal care services in the healthcare facilities. This study allows policy makers to make appropriate decisions to reduce neonatal and infant mortality in Bangladesh and help to achieve SDG in time.

196

EM-094 Online Shopping: A Case Study on Rajshahi University Students 1Md. Mahfuzur Rahman, 1Yasin Mollah, 1Shirajum Monira, 1Farzana Yesmin and 2Md. Monimul Huq 1 4th year honours student, Department of Statistics, University of Rajshahi, Rajshahi, Bangladesh. 2 Professor, Department of Statistics, University of Rajshahi, Rajshahi, Bangladesh. Correspondence: [email protected]

Internet technology has impelled significant changes in many features of social life and culture as well as in shopping ethos. Online shopping is also started in Bangladesh. But it is a new term and experience for a buyer to go into an online shopping. People feel uncomfortable and time consuming for going crowded markets. So, E-Shopping has advantage as it keeps lot of time. Web-based shopping is a method whereby customers directly buy goods, services etc. from a seller without an intermediary service over the internet. Shoppers can visit web stores from the comfort of their house and shop as by sitting in front of the computer. The objective of this study was to determine the level of acceptance towards online shopping and to identify the factors influencing online shopping of Rajshahi University students. A cross-sectional study was conducted and multi-stage simple random sampling technique was used for data collection. Different statistical techniques such as univariate, bivariate and binary logistic regression have been used in this study. From our study about 43% students engaged with online shopping while most of the students agreed that online shopping is more convenient than traditional shopping. But few students also think that traditional shopping is more enjoyable than online shopping. To understand the acceptance level of online shopping among the Rajshahi University students, we used Likert scale questions. Thirteen questions were asked to the students and saw the responses by measuring through mean and standard deviation. Besides this, the study highlighted that there is a significant relationship of online shopping with gender, duration of internet use per day, payment system and product price. On the other hand there is no significant relationship with income, delivery cost and delivery time. Web-based shopping is in a challenge with traditional shopping. At present web-based selling is the most immense and standard business area. Students of Rajshahi University are not habituated yet to go to an online shopping though a lot of online shops are developed. Online shops need some guidelines on trust on quality, secure payment gateway, prompt delivery system, deliver return policy to fulfill the need of the customers.

EM-095 Working in Volunteer Organization: A Case Study of Rajshahi University Students 1Rubina Akter, 1Yasin Mollah, 1Shirajum Monira, 1Most. Farjana Haque, and 2Md. Monimul Huq 1 4th year honours student, Department of Statistics, University of Rajshahi, Rajshahi, Bangladesh. 2 Professor, Department of Statistics, University of Rajshahi, Rajshahi, Bangladesh. Correspondence: [email protected]

Academic performance is the main identity of a student. A student verified his performance by their academic results. But now-a-days a student is not only judge by their academic performance but have some other skills. Student is judged by their result as well as their efficiency in presentation ability, various skills and innovative mind and one of the ways acquired these types of attributes through working on volunteer organization. So, the objective of this study was to investigate whether working in volunteer organization helps to improve students’ quality like their presentation ability, communication skill, organizing and management ability etc. From our study we observed that the percentage of male and female students was 89.84% and 10.16% respectively. We saw that there were no effects on results for working in volunteering organization. We also see that only 6.25% students are used text book for their study, 17.97% students are used hand note and 75.78% students are used both text book & hand note for their study. We also see about 30.5% students are doing library work and 69.5% student doesn’t. In cross table we saw that results were related to daily study hour and for working in volunteer organization don’t cause any obstacle to their study. We think students who are working in volunteer organization are more active than others. They have been performing a great job in social work and other extra curriculum activities. Working in volunteer organization they have built a good personality which is important for future career as well as academic results. The administration should remain alert to maintain a fair environment for the students that they could improve their ability and serve the nation efficiently.

197

EM-096 OBSTACLES IN THE WORKPLACE OF MAIDSERVANTS AND SOCIAL ATTITUDES TOWARDS THEM: A STUDY ON RAJSHAHI CITY CORPORATION Mst. Rushama Afroz1, Golam Kibria Ferdous1 and Ezaj Ahmad2 1 Professor, Department of Social Work, University of Rajshahi, Bangladesh 2 Department of Arabic, University of Rajshahi, Bangladesh

Bangladesh is an unprivileged and developing country of third world. Here poverty and employment are the most crucial problems ever existed. Because of extreme poverty and lack of unemployment many women in this country are assigning themselves as maidservants from ages and being molested by their employers’. Maid servants as part of the unorganized labor are today victims of unmitigated hardships and indignities. They are paid shockingly low wages. There are no minimum wages fixed as the Minimum Wages Act does not cover the domestic workers. They do not have any job protection or security as they are hired and fired at the will and fancies of the employers. For even trivial reasons the maid servants can lose her their job if the employers desire so. In the absence of any suitable legal measures to protect the interests of domestic/maid servants their plight remains unchanged and they are victims of exploitation. A present estimate of ASK (Ain O Salish Kendra) shows that in 2018 the total number of violated and tortured maidservants are 54. Where the rate of physical torture is 26, rape is 2, attempt to rape is 1, death (not mentioned the type of torture) is 16, the rate of suicide is 5 and death after physical torture is 2. But unfortunately, within these only 33 case has been filed and among these 33 cases, there is no information on case filling rate is 21. Millions of words have already been spent touching upon the various problems of women working in organized field but a very few studies on the women working in unorganized sector and still fewer on maid servants. The present paper aims to take up the task of exploring the obstacles of maidservants especially in their working and living conditions and for diminishing these obstacles both government and nongovernmental initiatives should be taken to ensure their safety and security in their workplace and prominent campaign should be run for diminishing humiliating attitudes towards them from the society. Submission for student paper/poster competition? Yes

EM-097 Achievement in child health in SAARC region: Trend and future Sultana Rajia*, Md. Kamrul Islam*, Md. Sabiruzzaman**, Md. Jahanur Rahman* and Dulal Chandra Roy* * Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh ** Corresponding author: [email protected]

This study has examined some important child health indicators of Bangladesh and provides a comparison among SAARC countries. Yearly time series data from 1994 to 2018 on four indicators - neonatal mortality rate (NMR), under 5 mortality rate (MR-U5), infant mortality rate (IMR) (per 1,000 live births) and prevalence of anemia among children under 5 ANM-U5) have been collected from World Bank data archive. ARIMA models have been fitted and used to forecast these indicators up to 2030. This study revealed that despite good progress in child health in the South Asia, the present situation in some countries is found to be vulnerable. The mortality rates of children in all countries show decreasing trend. Maldives and Sri Lanka have already achieved SDG target of neonatal mortality rate below 12 and under-five mortality rate below 25 per 1,000 live births. Other countries except Pakistan and Afghanistan are expected to reach the SDG target by 2030. The prevalence of anemia is extremely high in this region. For instance, more than 50% of children in India and Pakistan suffer from anemia. Moreover, the recent trends in prevalence anemia in some countries like Afghanistan, Pakistan, Sri Lanka and Maldives are increasing over time. Findings of this study provide useful information for policy makers and leaders of SAARC region. Keywords: Child health, Neonatal mortality, Anemia, ARIMA model. Submission for Student Poster Competition? Yes

198

EM-098 Prevalence and determinants of internet addiction among young adolescence: A survey in Rajshahi city Md. Mehedi Hasan*, Md. Kamrul Islam, Md. Sabiruzzaman**, Farhana Hasan, and Md Golam Hossain * Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh ** Corresponding author: [email protected]

Internet addiction appears as a major problem in adolescents. Secondary school represents an important time point in adolescents’ life, when they start enjoying freedom of life and thus, vulnerable to uncontrolled internet use. The study provides important insights regarding students’ access, skills, preferences and appetite for internet use; prevalence of internet addiction and its potential factors have been determined. Data have been collected from 622 students of class IX and X of different disciplines and different types of schools in Rajshahi city under a project funded by Faculty of Science, Rajshahi University. It is evident from the study that although majority of the secondary level students use internet to some extent, not all of them are using it for study regularly. More than 11 percent of the students are found to be addicted to internet and the proportion of addiction is higher in boys than girls. Relationship with parents, time of internet use and internet use for academic purpose has significant impact on internet addiction. Findings of this study convey important messages for the guardians and the policy makers as well. Keywords: Internet addiction, Adolescents, Secondary level education, Logistic regression. Submission for Student Poster Competition? Yes

EM-099 Gender differences in internet use preferences among high school students: A survey in Rajshahi city Mst. Sharmin Aktar Mukta*, Md. Sezon Islam, Md. Ismail Haque, Farhana Hasan, and Md. Sabiruzzaman** * Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh ** Corresponding author: [email protected]

The internet use by students has increased dramatically in recent years and also surfing the internet has become an important part of general student life. To improve standard of education with diversified opportunities in internet, ICT education is already in place in Bangladesh. In this context, the authors’ survey aimed at identifying gender differences in internet access and use preferences among secondary level students in Rajshahi city. Data have been collected from 622 students of class IX and X of different disciplines and different types of schools in Rajshahi city under a project funded by Faculty of Science, Rajshahi University. Gender difference in the pattern of internet use among high school students is evident from this study. Interests of students on various streams of internet are found to be differing significantly between boys and girls. While greater proportions of boys prefer using internet for chatting or Facebooking, more of girls are using internet for study purpose. Findings of the study suggest that girls are more functional than boys in internet use for academic reason. Keywords: Internet use, Gender difference, Secondary level education. Submission for Student Poster Competition? Yes EM-100 Levels and trends in greenhouse gas emission in SAARC countries Md. Roky Rahman*, Md. Kamrul Islam, Md. Sabiruzzaman**, and M. Sayedur Rahman * Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh ** Corresponding author: [email protected]

Greenhouse gas (GHG) is primarily responsible for global warming that includes rapid increase of globe temperature and climate change. Carbon dioxide (CO2) is one of the major GHG and is produced in cumulative amount with increasing industrialization and dwindling forestation. This study portrays the past and present scenario of GHG release in SAARC countries and focuses on future direction. Yearly time series data on CO2 emission, total GHG emission and proportion forest area in these countries from 1990 to present have been collected from World Bank data archive. It is evident that over the last three decades, emission of GHG in all eight countries has been increased to some extent. In 2014, highest CO2 emission is recorded in Maldives (3.07 metric ton/capita) followed by India (1.73 metric ton/capita). On the other hand, the proportion of forest area remains unchanged in Maldives and Afghanistan. It has been increased a little in only two countries - India and Bhutan. In other countries Bangladesh, Pakistan, Nepal and Sri Lanka,

199

proportion of forest area has been decreased. Only Bhutan, Sri Lanka and Nepal have forests more than 25% of their total lands. As revealed from the trend analysis that if no control measure is taken, in near future, accumulated GHG will be at alarming state for sustainable environment causing sea-level rise from thermal expansion to ocean acidification. Since all countries share same atmosphere, climate change is not just a national concern. This study conveys useful findings for the leaders of SAARC countries and for the policymakers, active in the area of environment and climate change as well. Keywords: Greenhouse gas, Carbon dioxide emission, Forestation, Global warming, Sustainable environment Submission for Student Poster Competition? Yes

EM-101 TOWARDS PUTATIVE PREDICTIVE ALGORITHMS FOR SEVERE DENGUE Dr. Shamala Devi Sekaran, Filippo Bosco, Chandramathi Samudi Raju and Chong Zhou Lin. Professor & Director of Basic Medical Sciences, Faculty of Medicine, MAHSA University, 42610 Jenjarum, Selangor email :[email protected]

Dengue fever virus is an arthropod-borne RNA virus transmitted by infected mosquitoes. It is the leading cause of serious illness and death among children across the tropical and sub-tropical belt. It has rapidly expanded over the last 50 years in the absence of an effective licensed vaccine and vector control. Currently WHO has classified three groups: dengue without warning signs; Dengue with warning signs and Severe Dengue. Most cases are non-severe. However, some cases result in plasma leakage and can be fatal. It is critical to identify the patients who are at risk of severe disease and likely to benefit from hospital care. No existing diagnostic tool is perfect when considering availability, sensitivity, and functionality early in the disease progression. Our main objective is to construct and demonstrate a dengue diagnosis and classification algorithm, thereby improving the diagnostic accuracy of dengue cases beyond that which is currently seen in Malaysian clinics. A clinical decision support system is needed to increase the speed and accuracy of diagnoses. Relatively little work has been done on developing such systems despite the promise they hold for the treatment of dengue. Current focus has been more towards predicting outbreaks than on the decision the physician on duty has to make and the plight of the individual patient. Work on diagnostic algorithms is comparatively sparse, with much of the available work being simple application of basic machine learning algorithms to raw patient data. We used the standard process for data mining (CRISP-DM) methodology. First we compared data between all sites and then sought to understand the data by exploratory data analysis (EDA) through descriptive statistics and correlation analysis, as well as clustering and dimensionality-reduction methods (e.g. principal component analysis) aimed at identifying features with the most predictive power. This phase also includes the identification of variables; dependent (diagnosis, classification), independent (base features), confounding (those which affect independent variables), and background variables. Analysis conducted to date will be presented.

EM-102 An in depth investigation on the effect of low frequency glow discharge plasma (LFGD) on the improvement of seed germination rate and agronomic traits in maize 1 1 1 1 Sohanur Rahman Sohan , Mahedi Hasan , Forhad Hossain , Sumon karmoker Ahmad Humayun Kabir2, Mamunur Rashid Talukder3, Abu Reza1* 1 Molecular Biology and Protein Science Laboratory, Department of Genetic Engineering and Biotechnology, University of Rajshahi, Rajshahi-6205, Bangladesh 2 Department of Botany, University of Rajshahi, Rajshahi-6205, Bangladesh 3 Department of Electrical and Electronic Engineering, University of Rajshahi, Rajshahi-6205, Bangladesh

Plasma, which is also known as fourth state of matter is recently gaining increasing interest in agriculture and biomedical Science. Current study explored the effect of LFGD Argon-Air plasma treatment generated from Argon-Air gas on seed germination, seedling growth and antioxidant enzyme activity in maize (MIRACLE 300). Plasma was generated in the laboratory by gas to an extremely high voltage. Maize seeds were treated with LFGD Argon-Air plasma for 30 sec, 60 sec, 90 sec and 120 sec. The result showed that seed treated with plasma had 12-14% higher germination rate compared to controls. Characteristics of morpho physiological parameters including plant height, stem diameter, fresh and dry weight of leaves and roots, chlorophyll concentration were also improved significantly by plasma treatment. Biochemical analysis

200

demonstrated that the H2O2 and NO concentration was increase substantially in leaf and roots of treated plants. Two important enzymes namely ascorbate peroxidase (APX) and superoxide dismutase (SOD) activities along with increase in antioxidant enzyme activities were found in leaf and roots. Thus, the present study revealed that LFGD Argon-Air plasma is involved in elevation of ROS in leaf and roots which in turn improves the seed germination rate and other agronomic traits.

EM-103 Enhancement of seed germination rate and seedling growth in maize using LPDBD plasma Mahedi Hasan1, Sohanur Rahman Sohan1, Forhad Hossain1, Musatim Billah1, Sumon karmoker1, Ahmad Humayun Kabir2, Mamunur Rashid Talukder3, Abu Reza1* 1Molecular Biology and Protein Science Laboratory, Department of Genetic Engineering & Biotechnology, University of Rajshahi, Rajshahi-6205, Bangladesh. 2Molecular Plant Physiology Laboratory, Department of Botany, University of Rajshahi, Rajshahi-6205, Bangladesh. 3Plasma Science and Technology Laboratory, Department of Applied Physics and Electrical Engineering, University of Rajshahi, Rajshahi-6205, Bangladesh.

This study investigates the effect and mechanisms of low pressure dielectric barrier discharge (LPDBD) plasma produced from argon (Ar) and oxygen (O2) gas which causes biological stimulation leading to improved germination and growth in Maize. Both plasma treatments caused noticeable improvement in seed germination in Maize. Beside this, root and leaf H2O2 concentration significantly increased compared to controls subjected to (Ar+O2). Analysis of plants grown from the plasma treated seeds showed significant improvement in shoot characteristics, stem diameter, fresh weight, dry weight and chlorophyll content in comparison with the controls. Interestingly, (Ar+O2) treated plants showed a significant increase of H2O2 and NO compared to controls in both root and leaf. This phenomenon was supported by the biochemical and molecular evidence of SOD and CAT in Maize plants. Plants derived from (Ar+O2) treated seeds demonstrated a significant increase in SOD and CAT expression in roots and leaf of Maize. This suggests that (Ar+O2) plasma caused a slight induction in H2O2 and NO accumulation without triggering their scavengers (SOD and CAT) and thus, efficiency affect growth and development in Maize plants.

EM-104 Statistics and SDGs: A Nexus Dipankar Roy, PhD Deputy Secretary and Focal Point (SDGs), Statistics and Informatics Division (SID), Ministry of Planning [email protected]

Monitoring the progress and attainment of Sustainable Development Goals (SDGs) is key to ensuring the right path of sustainable development. Data disaggregation is of great importance in monitoring the progress of SDGs particularly for ensuring the pledge that nobody will be left behind. In the traditional survey, sampling plan, by design, fails to provide the estimation at finer level of domain. Hence, small area estimation technique can be a big tool in producing estimation at any disaggregation level subject to the level of precision. Official statistics generated by the national statistical system including the national statistical office is not enough for monitoring the progress of SDGs. As such, there is a need for integrating and mainstreaming administrative data into official statistics. It is evident that, traditional sources are not enough in monitoring the progress of SDGs. Beyond the traditional sources, big data can be a modern source to be integrated into official statistics. Handling big data requires high skill and analytical capacity. Data scientists can easily deal with issues relating to data analytics. In the era of SDGs, statisticians should be the data scientists and storyteller. Thus, the paper focuses the role of statisticians and tries to make a nexus between statistics and SDGs. Finally, it argues that, sustainable development can be monitored by the sustainable statistics.

EM-105 Influence of sex preference on Demand for a Child in Bangladesh Saroje Kumar Sarkar1* and Md. Abul Basher Mian1 1 Department of Statistics, Rajshahi University, Rajshahi, Bangladesh. * E-MAIL: [email protected]

If son preference were to affect the demand for additional children then, within any parity, those with one or more sons would be more likely not to want more children as compared to those who did not have any sons. Conversely, if the desire for a balance sex composition was to affect fertility within a given parity,

201

respondent who had either all sons or all daughters would be more likely to want additional children as compared to those who had children of both sexes. Data from BDHS 1999-2000 clearly indicates a significant evidence of son preference in Bangladesh. The findings indicate that whenever gender preferences would be eliminated, the percentage of respondents who did not want more children would be increased approximately by 10 percent. Finally, we conclude that further reduction of fertility to achieve replacement level is unlikely without considerable reduction in the desire of male children or sex preferences.

EM-106 Recent Advancement of Growth Models and Longitudinal Data Md. Ayub Ali1 1Professor, Department of Statistics, University of Rajshahi, Rajshahi- 6205, Bangladesh Email: [email protected]

The purpose of the present study was to highlight the recent advancement of growth models and longitudinal data in Bangladesh. A review of growth models together with the innovation of new models and generating longitudinal morphological data will enhance our medical sector through creating different growth charts and fitting growth models. This study will help the researchers as well as the health sector of the government. Keywords: Growth Models, Longitudinal Data, Growth Chart

EM-107 Data Gap in SDGs: Bangladesh Perspective Md. Ayub Ali1 1Professor, Department of Statistics, University of Rajshahi, Rajshahi- 6205, Bangladesh Email: [email protected]

The purpose of the present study was to highlight the unavailability of research area under different SDGs in Bangladesh. A critical analysis show that many tires of the government have been being engaged to fulfill the achievement of SDGs and many areas are still remain to be undone. This study is showing those areas so that one can uphold his expertise for the development of our country. This study will help the researchers as well as the government. Keywords: Data Gap, SDGs, Action Plan

EM-108 Meta-Analysis of Age Related Macular Degeneration using Fixed, Random and Mixed Effects Models Suresh K. Sharma, Savita Jain and Kanchan Jain Department of Statistics, Panjab University, Chandigarh-160014 (INDIA) [email protected]

Meta-Analysis is the use of statistical methods for combining results of individual studies. This allows one to make the best use of all the information collected in a systematic review by increasing the power of the analysis. By statistically combining the results of similar studies, the precision of estimates of treatment effect can be improved and it can be assessed whether treatment effects are same in similar situations. Age- related macular degeneration (AMD) is a common degenerative disease among the elderly population, leading to distorted central vision in the early stage and severe visual loss in the advanced stage. AMD is a complex disease with multiple genetic and environmental factors. We analysis the data set with the results from 20 studies on the effectiveness of the treatment against AMD disease using fixed, random and mixed effect models. The heterogeneity of various studies has been tested suing Q-Statistic. The treatment is statistically significant for all the models. The analysis includes estimate of effect size and relative risk of AMD in the treated vs. control groups with the incorporation of 95% confidence intervals for all the three models. Various graphs including Forest, Funnel, Radial, QQ are also plotted for fixed, random and mixed effects models. For mixed effects model, only year moderator appears to have a significant influence on the effectiveness of the treatment although based on the QM test; both moderators have significant influence on the treatment. Keywords: AMD, meta-analysis; fixed-effects model; random-effects model; mixed-effects model.

202

EM-109 Impact of Demography on Health: A Study on Three Decadal Censuses of India Dr. Palash Chandra Coomar1, Purnendu Kishore Banerjee2, Ricklina Roychowdhury3 1 Former Joint Director of Census Operations, Ministry of Home Affairs, Govt. of India 2 Director, Central Statistical Organization, Ministry of Programme Implementation, Govt. of India 3 Junior Research Fellow, Anthropological Survey of India, Ministry of Culture, Govt. of India

Health may be defined as a complete physical, social and mental well being and not merely absence of disease or infirmity. Our physical, mental and social well being are precondition by demographic parameters. Disequilibrium in population dynamics, sex ratio, disparity in sex specific migration, rural- urban population, State/ Union Territory, different social groups i.e. Scheduled Caste and Scheduled Tribes, religious groups wise population profile is highly influencing availability of suitable marriage partner, social security of the vulnerable segment of population and all these conjointly independently in turn have decisive impact on physical, social and mental well being of health of People of India. In the present paper attempt has been made to show State / Union Territory wise sex ratio, rural – urban sex ratio, sex ratio of different social and religious groups of India as revealed from last three decadal censuses i.e. Census of India, 1991, 2001 and 2011. To show the impact of demography on health of population in India.

EM-111 RNA-seq Data Analysis to Identify Cancer Related Biomarker Genes Md. Ariful Islam1*, Md. Emdadul Haque1, S. Amena Tasmia1, Md. Hadiul Kabir1 and M. Nurul Haque Mollah1 1Bioinformatics Lab., Dept. of Statistics, Rajshahi University, Bangladesh *Poster presenting author ([email protected])

RNA-seq data has an increasing popularity over traditional microarrays in transcriptome analyses. Statistical methods used for gene expression analyses are different because the array-based technology measures intensities using continuous distributions, whereas RNA-seq provides absolute quantification of gene expression using counts of reads. RNA-seq offers several advantages over microarrays for differential expression analysis, such as an increased dynamic range and a lower background level, and the ability to detect and quantify the expression of previously unknown transcripts and isoforms. The analysis of RNA-seq data is, however, not without difficulties. In differential expression analysis, where the genes are tested individually for expression differences between conditions, such ‘within-sample’ biases are usually ignored since they are assumed to affect all samples similarly. There are several methods for identification of DE genes from RNA-seq datasets. Some methods work directly on the count data, while the remaining methods transform the counts and feed the transformed values into the R package LIMMA, which was originally developed for differential expression analysis of microarray data. The methods working directly on the count data can be broadly divided into parametric (baySeq, EBSeq, ShrinkSeq, edgeR, DESeq, NBPSeq, and TSPM and non-parametric methods (NOISeq and SAMseq. In this work, we would like to explore better statistical approaches for RNA-seq data analysis. The performance of the statistical approaches was investigated using both synthetic and real RNA sequence analysis. Keywords: RNA-seq, Count data, Differentially expressed gene, Cancer diseases, Gene ontology,

EM-112 Deep Learning for Identification of Rare Cell Types from Single-Cell Transcriptomic Dataset Md. Emdadul Haque1*, Md. Ariful Islam1, Atul Chandra Sinha2 Md. Mostafizur Rahman1 and Md. Nurul Haque Mollah1 1 Bioinformatics Lab., Dept. of Statistics, Rajshahi University, Bangladesh 2Department of Statistics, Begum Rokeya University, Rangpur, Bangladesh *Poster Presenting Author ([email protected])

Single-cell RNASeq (scRNASeq) has emerged as a powerful method for quantifying the transcriptome of individual cells at low cost and high throughput. In this era, it is most attractive and promising research field in bioinformatics. There is substantial ongoing effort to use scRNAseq measurements to identify the “cell types” that form components of a complex tissue in biological science. However, the data from scRNAseq

203

experiments is often both noisy and high dimensional, making the computational analysis non-trivial. Cell type identification from scRNAseq data involves the application of computational tools rooted in clustering and statistical analysis to identify molecular signatures that are unique to each type. Since the datasets continue to grow in size and complexity, then the demand for computational methods are also become abound for the analysis of scRNAseq data as scalable, flexible, and robust. Recently, huge growth in single cells it is expected to find the rare/new cell types in scRNAseq analysis. Although some tools have been developed for recognizing cell clusters based on the scRNAseq data, as far as we know, systematic cell type identification methods that sort cells into different rare or new groups/states have not been developed. Therefore, in this study we would like to propose a deep learning method that is scalable, flexible and robust for identifying rare/new/unknown cell types. Keywords: Single-cell RNASeq (scRNASeq), Cell type, Rare cell type, Cell type Identification.

EM-113 Statistical Genome-wide Association Studies to Detect Important SNPs Md. Golam Firoj*, Md. Mehedi Hasan, Md. Kaderi Kibria Md. Mamun Monir and Md. Nurul Haque Mollah 1 Bioinformatics Lab., Dept. of Statistics, Rajshahi University, Bangladesh *Poster Presenting Author ([email protected])

A single nucleotide polymorphism (SNP) is a variation at a single position in a DNA sequence among individuals. It is a substitution of a single nucleotide that occurs at a specific position in the genome, where each variation is present at a level of more than 1% in the population. SNPs make up over 90% of all human genetic variations, risk to certain diseases and response to drugs. SNPs underline differences in our susceptibility to a wide range of diseases (e.g. Cancer, Sickle-cell anemia, β-thalassemia, Cystic fibrosis, etc.). There are more than1000000 SNPs for the human population. Among these huge candidates, few SNPs say 1-20 SNPs are usually significantly associated with the occurrence of a certain diseases or any other phenotypic variations. Identification of these few SNPs responsible for a certain diseases or any other phenotypic variation out of huge number of candidate SNPs is a challenging task.To explore better statistical algorithm for identification of disease related important SNPs, we used both simulated and real SNP datasets, where the real datasets were downloaded from the online databases (Swiss-Port, NCBI, dbSNP, dsSNP, GDB, EMBL). In this study, we considered contingency table, linear regression, ridge regression, linear mixed model, etc. Also some performance measures were used to assess the performance of different methods method, such as FDR and power curve, etc. We considered some computer software (EMMA, plink, GenABEL, etc) during the comparative study. We observed that mixed model approach outperform over the other approaches to detect important SNPs. Keywords: Single nucleotide polymorphisms (SNP), Genome-wide association studies (GWAS), Linear mixed model, Some computer software, Performance measure.

EM-114 De-Novo Assembling for NGS Data Analysis to Detect Diseases Related Microbes Md. Kaderi Kibria1*, Md. Mehedi Hasan1, Md. Golam Firoj1 Md. Mazharul Islam1, Md. Mamun Monir1 and Md. Nurul Haque Mollah1 1Bioinformatics Lab., Dept. of Statistics, Rajshahi University, Bangladesh *Poster presenter ([email protected])

Human body is the home of trillions of bacteria, viruses, fungi, and other tiny unicellular organisms. Most of the microbes in our body work in favor of our good health. There exist few microbes in the human body or in the biosphere responsible for a certain human diseases. When a diseases related microbes get favorable environment they become active in the human body and develop the diseases. Diseases related microbial information’s are used to develop new drug or vaccine. So without detecting the certain disease related microbes, it would difficult to develop an appropriate drug or vaccine for a certain disease. However, identification and characterization of few disease related microbes out of huge number of microbes can’t be readily achieved due to the limitations in culturing methods. To overcome the limitation of the culturing methods, recently researchers are using the microbial sequences generated by the next generation sequencing (NGS) based most advanced technology. NGS allows massively parallel sequencing with thousands to millions of short sequence fragments in one experiment at considerably low cost compared to other sequencing technologies. It requires sequence assembling which refers to aligning and merging sequence fragments from a longer DNA sequence in order to reconstruct the original sequence, since there is no sequencer yet to produce full nucleotide sequence of a genome (DNA/RNA). De-novo sequence assemblers

204

are a type of algorithm that assembles short nucleotide sequences into longer ones without the use of a reference genome. Different assemblers are designed for different type of read technologies. Reads from second generation technologies (called short read technologies) like Illumina are typically short (with lengths of the order of 50-200 base pairs). However reads from third generation technologies like PacBio and fourth generation technologies like Oxford Nanopore (called long read technologies) are longer with read lengths typically in the thousands or tens of thousands. This necessitates different algorithms for assembly from short and long read technologies. In this work, an attempt is made to investigate the performance of different de-novo assemblers (SPAdes, Ray, ABySS, ALLPATHS-LG, Trinity, HGAP, Falcon, Canu, Hinge, Velvet, etc.) to recover whole genome from NGS datasets highlighting the statistical contributions. Keywords: DNA sequence, Next generation sequencing, De-novo assembling, Computational algorithm and online software.

EM-115 A Comparative Study on Reference based Assembling for NGS Based Metagenomic Data Analysis Md. Mehedi Hasan1*, Md. Kaderi Kibria1, Md. Golam Firoj1 Md. Mamun Monir1 and Md. Nurul Haque Mollah1 1 Bioinformatics Lab., Department of Statistics, Rajshahi University, Bangladesh *Poster Presenter ([email protected])

Metagenomics is the genomic analysis of micro-organisms by direct extraction and cloning of DNA from an assemblage of micro-organisms. It is a modern genomics technique which combines statistics, microbial sequence and computer technology to explore important microbes/microbial-gene associated with the complex disease/traits. Next-generation sequencing (NGS) is a high-throughput methodology that enables rapid sequencing of the base pairs in DNA samples. It requires sequence assembling which refers to aligning and merging sequence fragments from a longer DNA sequence in order to reconstruct the original sequence, since there is no sequencer yet to produce full nucleotide sequence of a genome. A reference genome is the initial sequence to which all subsequent sequences are ultimately compared. Assembling the billions of nucleotide bases that make up the conifer genome is a huge challenge, but important because the reference genome leads to the identification of all or most of the genes in an organism, and reveals features of the genome structure, such as the amount and order of repetitive elements and the nature of regulatory elements. Re-sequencing, or sequencing, other individuals within the same species is vastly less time consuming and costly once a reference genome exists. There are several reference based genome assemblers. In this work, an attempt is made to investigate the performance of different algorithms for reference based assembling (Fastp, FastQC, Velvet, Ragout, BWA, Bowtie2, VarScan, SAMtools and BCFtools, etc.) to recover whole genome from NGS datasets highlighting the statistical importance in this area. Keywords: Whole genome sequence, Next generation sequencing (NGS), Short sequences, Reference sequence, assembling algorithm & Online software.

EM-116 Phylogenetic Clustering of Microbial Communities Based on 16S rRNA Sequences Md. Mazharul Islam*, Mst. Sharmin Akter, Md. Nurul Haque Mollah* Laboratory of Bioinformatics, Department of Statistics, University of Rajshahi, Rajshahi 6205, Bangladesh. * Presenting author ([email protected])

The 16S rRNA sequences are commonly using for identification, classification and quantitation of microbes in complex biological mixtures such as environmental samples like water, soil or air, and microbiome samples. The 16S rRNA sequences are also using for phylogenetic studies as it is highly conserved between different species of microbes. Phylogenetic clustering of microbial communities based on 16S rRNA sequence is playing the vital role to identify diseases related virus and bacteria. There are several methods for the formation of phylogenetic tree based on sequence data. Up to date, nobody compare the phylogenetic tree methods yet. Only few papers was using one of the phylogenetic tree methods among the several methods of phylogenetic tree. In this paper, an attempt is made to explore better phylogenetic tree approach for microbial clustering based on 16s RNA sequence. For microbial clustering we have used phylogenetic tree which has four methods (Neighbor-Joining, UPGMA, Maximum Parsimony and Maximum Likelihood) that have been used in this study for getting better result. The analyzing results of this study apprize that for microbial clustering, using phylogenetic tree based on sequence dataset, the maximum likelihood method is comparatively better than the other three methods. Keywords: Microbial clustering, 16S rRNA sequence, Phylogenetic tree, Maximum Likelihood approach. Diseases related microbes.

205

EM-117 Identification of potential therapeutic targets and agents of non-small cell lung cancer through bioinformatics analysis Md. Parvez Mosharaf* and Md. Nurul Haque Mollah Bioinformatics lab, Department of Statistics, University of Rajshahi, Bangladesh *Presenting Author ([email protected])

Lung cancer is the most important health risk for human in worldwide. Non-small cell lung cancer (NSCLC) is the most common cause of premature death from malignant disease. As like as other cancer disease, early diagnosis of NSCLC is essential which can significantly reduce the deaths rate. The current study focused to identify the differentially expressed hub genes and the molecular mechanism of the genes in NSCLC through the bioinformatics analysis. The total 1896 differentially expressed (DE) genes were identified including 1366 up regulated and 576 down regulated genes using expression data GSE19804 through GE02R server which consists of LIMMA. The protein-protein interaction network analysis was conducted and it revealed the hub proteins (CDK1, EGFR, FYN, UBC, MYC, CCNB1, FOS, RHOB, CDC6, CDC20, and CHEK1) which are mostly connected with rest of the proteins. The hub gene’s prognostic index was observed by using SurvExpress which reflected that the genes have a significant impact in normal and cancer patients. These hub proteins can be treated as the therapeutic targets for NSCLC. The molecular signatures and regulatory elements presented here permit further attention for experimental studies which are offer significant potential as biomarkers and candidate therapeutics for precision medicine approaches to clinical management of NSCLC. Keywords: Non-small cell lung cancer, Molecular signature, differentially expressed genes, Protein-protein interaction, bioinformatics approach. Submission for Student Paper Competition: Yes

EM-118 Factors associated with declining trends of HIV prevalence in India Dr. Pushpanjali Swain Professor and Head, Dept. of Statistics and Demography National Institute of Health & Family Welfare, New Delhi - 110067 India Email: [email protected]; [email protected]

Context: Being signatory to reduce HIV infections in India by 2030 to meet sustainable development goals 3.3, India is committed to end HIV epidemic in the country. HIV epidemic in India is concentrated in nature. It is heterogeneous among the different group of sub-population and also in the geographical distribution and the drivers of the epidemic is also different in the different geographical area of the country. According to recent round of HIV sentinel surveillance conducted in 2017, the overall adult HIV prevalence is 0.22% in the country, and the prevalence among Injecting Drug Users is high i.e. 6.26% followed by Transgender (3.14%), Men having Sex with Men (4.40%) and Female Sex Workers (1.56%). Methodology: The study uses data from various rounds of sentinel surveillance in the country. In the sentinel surveillance, India’s HIV epidemic pattern is monitored among the high-risk groups, bridge populations as well as the general population & provides transmission dynamics. Results: The HIV epidemic has a declining trend in the country as annual new infection has decreased by 27% as compared to year 2010. In 2017, around 87,580 new HIV infections cases estimated in the country. The HIV prevalence is decreasing among general adult population and all high risk groups except Intravenous Drug Users group. The programmatic interventions and continuous monitoring through surveillance helped to decline HIV prevalence in the country. For example, knowledge of HIV/AIDS has increased more among women than among men and knowledge has been continuously increasing over the years with education and wealth quintile. Conclusion: With scaling up ART centres in the country for the people living with HIV and package of preventive services to the community, country is committed to end from HIV epidemic by 2030.

206

EM-119 Community acquired multi-drug resistant clinical strains from tracheal aspirates of patients in hospital settings in Dhaka, Bangladesh. Fatema Tuz Jubyda1, Md Murshed Hasan Sarkar1, Md Ibrahim1, Tasmia Farzana2, Jinia Afroz2, Sanzida Sharmin2, Md. Jobaid Faruq2, Amit Kumar Dey1, Sabbir Ahmed1, Md. Zamilur Rahman1, Arfatun Nahar Chowdhury1, Md. Moniruzzaman1 1Bangladesh Council of Scientific and Industrial Research Laboratories, Rajshahi, Bangladesh. 2Primeasia University, Dhaka, Bangladesh Presenting Author: Fatema Tuz Jubyda ([email protected])

Abstract Antimicrobial resistance is a multi-sectoral problem which poses a major threat in the treatment of infectious diseases especially in developing countries like Bangladesh. Multidrug-resistant (MDR) bacteria along with extremely drug resistant (XDR) bacteria have emerged as major clinical and therapeutic dilemma in the treatment of tracheal infections in hospitals here. In this study, of 200 clinical, non-duplicate bacterial isolates obtained, Pseudomnas aeruginosa was the most frequent pathogens (N=61/200, 30.5%) followed by Acinetobacter baumannii (N=58/200, 29%), Klebsiella pneumoniae (N=45/200, 22.5%), Streptococcus pneumoniae (N=15/200, 7.5%), Escherichia coli (N=10/200, 5%), Staphylococcus aureus (N=4/200, 2%), Proteus spp (N=3/200, 1.5%), Enterobacter spp (N=2/100, 1%), Citrobacter spp (1/200, 0.5%), Providencia spp (N=1/200, 0.5%). Of 20 different antibiotics tested, highest number of isolates (N=172/200, 86%) showed resistance to third generation cephalosporin cefixime, however least number of isolates showed resistance to polymixin antibiotics- colistin (N=25/200, 12.5%) and polymixinB (N=12/200, 6%). The patients’ ages ranged between 1 month to 95 years with the gender distribution of 133 (66.5%) males and 67 (33.5%) females. The prevalence of infections was highest among the patients of age-group (old adults) ≥60 years (N=123/200, 61.5%). Of 200 clinical isolates, 43 (21.5%) were XDR and 125 (62.5%) were MDR bacteria. Of 200 clinical isolates, the synthesis of extended spectrum β-lactamases (ESBL) and carbepenemase was detected in 59 (29.5%) and 98 (49%) strains respectively. Therefore, there is an urgent need for constant surveillance and interventions in Bangladesh in order to prevent further spreading of those resistant organisms.

EM-120 Time Series Modeling and Forecasting of Somali Population Aweis Abdullahi Hersi1, Isse Osman Adem1, Md Abdul Khalek2 and Md. Ayub Ali3 1 MSc Student, Department of Statistics, University of Rajshahi 2Associate Professor, Department of Statistics, University of Rajshahi 3Professor, Department of Statistics, University of Rajshahi Corresponding Email: [email protected]

The present study aimed to build the time series model on Somali population and to forecast this population up to 2030. Data of population from 1960 to 2017 were collected from the World Bank (https://data.worldbank.org). Test of Stationarity was performed and it was found that it was stationary at 2nd difference and best fitted model for was ARIMA (1, 2, 1) with GARCH (0, 3). The forecasted population of Somalia will be 22121071 by 2030 out of which 10980359 will be male and 11140712 will be female. The Government should take necessary action on this result for the development of the country as a whole. Keywords: Time Series, Forecasting, Somali Population

EM-121 Time series Analysis modeling and Forecasting of Infant mortality in Somalia Isse Osman Adem1, Aweis Abdullahi Hersi1, Md. Abdul Khalek2 and Md. Ayub Ali3 1 MSc Student, Department of Statistics, University of Rajshahi 2Associate Professor, Department of Statistics, University of Rajshahi 3Professor, Department of Statistics, University of Rajshahi Corresponding Email: [email protected]

The present study purposed to build the time series model and forecast on infant mortality up to 2030. Data of infant mortality from 1983 to 2017 were collected from the World Bank (https://data.worldbank.org).

207

The fitted model was found as Yt = + β Yt­1 + Yt­2 +ɛt.. The fitted model was checked and validated accordingly and then the forecasted value of the infant mortality were found during the years 2018-2030. It is found that the infant mortality was𝛼𝛼 found to be 𝛿𝛿554900 by 2030. The government should take proper action to reduce this mortality. Keywords: Time Series, Forecasting, Infant Mortality, Somalia

EM-122 Stochastic Model on Fishery in Randomly Varying Environment M. A. Shah Department of Statistics, University of Rajshahi, Bangladesh

Estimation of population size and its response to harvesting, projection of future success, and selection of an appropriate strategy so as to maintain a stable sustainable yield (SSY) as well as to preserve a healthy population, are the most challenging and foremost problems of managing fisheries. The problem of management fisheries is complicated by the factor that the natural populations have a tendency to fluctuate in response to stochastic perturbations in their physical and/or biological environment. Population’s sensitivity to such random perturbations depends on fertility, mortality and recruitment and in general modified by harvesting. These effects become alarmingly severe if harvesting inadvertently exceeds the maximum sustainable yield (MSY). High efforts may result into low average yield with high variance. Keywords: Logistic model, Stochastic Model, Multiplicative fluctuations, Fish harvesting.

EM-123 Halting and Reversing Land Degradation and Unplanned Land-use under SDG-15 in Bangladesh Md. Sifat Hossain1, Md. Abdullah Al Mamun1, Md Abdul Khalek2 and Md. Ayub Ali3 1 BSc Student, Department of Statistics, University of Rajshahi 2Associate Professor, Department of Statistics, University of Rajshahi 3Professor, Department of Statistics, University of Rajshahi

Land degradation is a serious problem for achieving food securities (SDG-2) which may be occurred due to the rapid horizontal increasing of urbanization and unplanned land use. Removing this problem under SDG- 15 is a challenge now-a-days because the reality is yet to be investigated and pointed it out. The present study highlighted the pin-point of the problem of losing agricultural land and possible robust solution to overcome it. A declining trend was observed for the total agricultural lands of the country, i.e. a decrease is noted from 91.83% in 1976 to 87.69% and 83.53% over the years of 2000 and 2010 respectively. A total of 561,380ha agricultural lands were decreased during 1976-2000 and this figure was increased to 565,370ha during 2000-2010. Yearly average loss of agriculture lands were 23,391ha and 56,537ha during 1976-2000 and 2000-2010 respectively. This indicates that agriculture lands were transforming to other activities in higher rates between the years of 2000 and 2010. This is definitely a threat to achieve SDG-2 and SDG-15. Present study highlighted a possible way to solve the problem / threat. The government as well as the proper authority should take necessary action on this result for the development of the country as a whole. Keywords: Land Degradation, Agricultural Land Use, Robust Solution, SDG-15, SDG-2

EM-124 Marine Resources: Challenges, Opportunities and Realities Md. Abdullah Al Mamun1, Most. Israt Jahan1, Md Abdul Khalek2 and Md. Ayub Ali3 1 BSc Honors Student, Department of Statistics, University of Rajshahi 2Associate Professor, Department of Statistics, University of Rajshahi 3Professor, Department of Statistics, University of Rajshahi Corresponding Email: [email protected]

Getting of marine resource is decreasing day by day. Removing this problem under SDG-14 is a challenge now-a-days because the reality is yet to be investigated and pointed it out. The present study highlighted the pin-point of the problem of losing lives & wealth and possible robust solution to overcome it. The government as well as the proper authority should take necessary action on this result for the development of the country as a whole. Keywords: Marine Resource, Loss of life, Robust Solution, SDG-14

208

EM-125 Enhancing Access to Modern Energy under SDG-7 in Bangladesh Md. Abdullah Al Mamun1, Most. Israt Jahan1, Md Abdul Khalek2 and Md. Ayub Ali3 1 BSc Honors Student, Department of Statistics, University of Rajshahi 2Associate Professor, Department of Statistics, University of Rajshahi 3Professor, Department of Statistics, University of Rajshahi Corresponding Email: [email protected]

Different kinds of energies have been used in Bangladesh e.g., Electricity, coal, gas, oil, wind, water, and solar. The use of electricity is higher however it needs either coal or gas or oil as raw materials whose reservation is supposed to be decreased day by day. Removing this problem, it is needed to be enhanced the technological sector of generating electricity from water, wind and solar energies. The present study highlighted the pin-point of the cheapest energy. A comparative study among the cost-managements of different energies divulge that the solar energy is the cheapest and harmless. However, to lead a sophisticated life, we must have to ensure certain level of energy. The government as well as the proper authority should assign to the universities to take necessary action on this result for the technological development of the energy sector. Keywords: Energy, Electricity, gas, oil, water, wind, solar energy, technological development, SDG-7

EM-126 Productive Employment and Decent Work for Bangladeshi Expats under SDG-8 Most. Israt Jahan1, Md. Abdullah Al Mamun1, Md Abdul Khalek2 and Md. Ayub Ali3 1 BSc Honors Student, Department of Statistics, University of Rajshahi 2Associate Professor, Department of Statistics, University of Rajshahi 3Professor, Department of Statistics, University of Rajshahi Corresponding Email: [email protected]

Every year, Bangladesh is exporting manpower in different countries. Among the expats, about 80% are exported in the middle east countries where the language is Arabic. Due to this language and also lack of proper training, Bangladeshi workers as well as the country as a whole is getting less benefit/salary. The total amount of monetary loss per month only from Saudi Arabia, in this connection, is 1200000x40000=48,00,00,00,000Tk to 2100000x40000=8400,00,00,000Tk. Bangladesh is getting also similar loss from other countries. On the other hand, low graded employers are tyrannized and assaulted everyday that is a shame for the nation as a whole. Possible solution is to teach the language and train them. The government as well as the proper authority should take necessary action on this big issue. Keywords: Manpower Export, Expats, Lack of Training, SDG-8, Language

EM-127 Adequacy of Summary Measures of Fit in Binary Response Model S. K. Sarkar and, Habshah Midi

Logistic regression is frequently used in social science studies since last decade. In spite of such applications, there has been no commensurate increase in the use of commonly available methods for assessing the model adequacy. Failure to address model adequacy may lead to incorrect inferences. Few existing robust and informative summary measures are proposed to employ for assessing the fit of the model before proceed to present the results from a fitted model. The presented optimal summary measures of fit suggest that the fitted model is sufficiently adequate for prediction. The proposed modified likelihood ratio test presented in the study suggests that the excluded variables are not potential for the response variable. It is strongly recommended that the methods for assessing model adequacy in different aspects should be used before proceed to presenting the results from a fitted binary response model. Adequacy check of the model is essential for prediction and may be helpful to the policy maker for policy implications in social science studies.

209

EM-128 Comparison of the Performances of Machine-Learning and Classical Models for Forecasting Climatic Variables in Rajshahi Md. Syful Islam1, Most. Tawabunnahar2 and Md. Jahanur Rahman3 1 Ph.D. Fellow, Institute of Environment Science, University of Rajshahi, Rajshahi-6205, Bangladesh 2 Dept. of Statistics, Jatiya Kabi Kazi Nazrul Islam University 3 Professor, Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh

For the first time, an attempt has been made in this study to use two machine-learning methods (ANN and SVM) for modeling and forecasting five climatic variables in Rajshahi along with the commonly used classical models ARIMA. The climatic variables are maximum temperature, minimum temperature, rainfall, relative humidity and sea level pressure. We have fitted three models ARIMA, ANN and SVM for each climatic variable. Then the models are compared based on their forecasting accuracy in order to determine which one of these three models offers highest forecasting accuracy. To evaluate forecast accuracy three performance measures namely MSE, RMSE and MAPE have been used for each model. It has been shown that the ANN as the best model for all climatic variables in Rajshahi.

EM-129 Asymptotic Properties of Maximum Likelihood Estimators Using Warranty Data Consisting Only of Failure Information Md. Mesbahul Alam Department of Statistics, University of Rajshahi, Rajshahi-6205, Bangladesh E-mail: [email protected] Improving reliability of product is an important part of the larger overall picture of improving product quality. Warranty database is a preferred source of such knowledge which is automatically generated and updated at no additional cost from repair claims during warranty coverage. For engineering purposes, usage time (e.g., mileage or copy number) is more relevant and lifetime parameters measured in usage time is an integral part of reliability analysis using warranty data. The reliability analysis becomes complicated because warranty data generally consist of only failure-related information, and non-failure information is not included. To overcome this situation, Alam and Suzuki (2009) discussed the maximum likelihood estimation on usage time scale based on warranty data consisting only of failure information. They also investigated the properties of the estimators from observed Information matrix. For this purpose, they considered Weibull as lifetime distribution and Log-normal as censoring distribution. This study is an extension of the work of Alam and Suzuki (2009). For this purpose, the asymptotic properties of the estimators are investigated based on Fisher Information matrix and hence the expected values of the elements of the Hessian matrix are derived, and found that if one utilizes the failure information only, one can estimate the lifetime parameters as well as censoring parameters with good precision. Again, both the observational period and the number of sold products affect so much on the precision of the estimators. Generally, as the number of sales amount of automobile is so large, one can estimate the lifetime parameters with sufficient precision. Keywords: Warranty data, Reliability, Censoring, Lifetime distribution

EM-130 Development of Risk Classification for Stomach Colon Cancer Md. Al-Amin1, Khanis Farhana Tuly1, Al Amin1, Md. Shadikug Jaman1, Umma Hafsah Himu1, Most. Shamole Akter1, Md. Maeen Molla1, Mahfuza Akter1, Mst. Sabina Yeasmin Swapna1, Roksana Akter Ritu1, Md. Aminul Hoque1* 1 Department of Statistics University of Rajshahi, Rajshahi-6205 * Correspond Author: [email protected] No doubt cancer is the most dangerous killer disease among other threatening diseases. However, cancer status is also an important issue to be addressed because stage of cancer rather more important for cure up. Even early stage cancer is not so dangerous for life threatening. Therefore it is urgent to classify risk of cancer stages. We may identify the stomach cancer risk as low, intermediate and high. This study has revealed a general application of survival analysis, a popular data analysis approach for certain type of cancer clinical data. Here we focused on the problem addressed by survival analysis using cox hazard survival model, multiple logistic regression model and Kaplan Meyer model to identify disease risk classification for stomach cancer. We developed here a new technique to stomach cancer risk classification

210

for prognostic prediction and find the risk free probability for the patients at the specific period of time. We used 16496 stomach Japanese cancer patients of which 11364 (68.9%) was male and 5132 (31.1%) was female. About 63.1% of the total patients belongs to the age group 60 to 80 years and 25.6% was in the age group 40 to 60 years. Our new technique would certainly give good contribution health sector and medicine. Key Words: Survival Analysis, Risk classification, Development, New Technique, Hazard Model

EM-131 An Overview Of Hydrogen Fuel Cell Vehicles: Fundamentals, Advancements And Future Prospects Md. Shehan Habib 1* 1Scientific Officer, Bangladesh Council for Scientific and Industrial Research *Corresponding author e-mail: [email protected]

The detrimental effects of pollutants from conventional fuel vehicles have led to the need for environmentally friendly energy sources. Hydrogen can be the most promising source of renewable energy. Hydrogen fuel cell can produce electricity by allowing chemical gases and oxidants as reactants. The fuel cell splits the caution and the anion in the reactant with anodes and electrolytes to produce electricity. The entire technology is environment friendly and produce water as a by product. The fuel cell can produce direct current (DC) power to run the electric car. By integrating a hydrogen fuel cell with batteries and the control system with strategies, one can produce a sustainable hybrid vehicle. The benefits of hydrogen and fuel cells are numerous, but will not be fully apparent until they are in widespread use. Research for hydrogen fuel cell electric vehicles is important to develop technologies that advance fuel cell systems, subsystems, and components. . In this study, the fundamentals, recent development and future prospects are reviewed to analyze the practicability of consuming hydrogen as a main fuel in vehicles.

EM-132 Statistical Modeling and Nomograms for Predicting Overall Survival in Lung Cancer Patients Md. Maeen Molla1, Most. Shamole Akter1, Umma Hafsah Himu1,Mst. Sabina Yeasmin Swapna1, Khanis Farhana Tuly1, Al Amin1, Md. Shadikug Jaman1, Mahfuza Akter1, Roksana Akter Ritu1, Md. Al-Amin1, Md. Aminul Hoque1* 1 Department of Statistics University of Rajshahi, Rajshahi-6205 * Correspond Author: [email protected]

The present study was to develop nomograms for predicting lung using a cohort from the Surveillance, Epidemiology, and validate the performance of the nomograms internally and externally. These nomograms would provide clinicians with quantitative tools to assess patients’ prognosis, for better patients’ risk classification and clinical decision making in the setting of surgically resected ICC. Here we used Statistical models for analysis and nomograms development with 111 lung cancer patients. Nomograms for lung cancer clinical data were established based on the independent prognostic factors identified by COX regression models and multiple logistic regression models, respectively. The performance of the nomograms was validated internally and externally by using the concordance index (c-index), and calibration plot, and compared with other existing techniques by using c-index and decision curve analysis. There were about 56% male and 44% female patients with age greater than 50 years was 60% and below age 50 was 40%. The developed nomograms could predict lung cancer with good performance, which could be served as an effective tool for prognostic evaluation and individual treatment strategies in clinical practice. Keywords: Nomograms, quantitative tools, Predictions, Survival Analysis, Prognostic evaluation, c-index

EM-133 Electrodeposition of Chromium from deep eutectic based ionic liquid Md. Moynul Islam*1,2 and M. Rostom Ali1 1Department of Applied Chemistry and Chemical Engineering, University of Rajshahi, Rajshahi-6205, Bangladesh. 2Department of Chemistry, Bangladesh Army University of Engineering and Technology, Qadirabad Cantonment, Natore-6431, Bangladesh.

The electrodeposition of chromium from a solution containing chromium chloride (CrCl3.6H2O) in choline chloride-ethylene glycol (EG) based ionic liquid has been carried out onto a copper cathode by constant current and constant potential methods at different temperature. The influences of various experimental

211

conditions on electrodeposition and the morphology of the deposited layers have been investigated by scanning electron microscopy (SEM) and X-ray diffraction (XRD). It is found that bright metallic coloured chromium coatings has been obtained from choline chloride-EG based ionic liquid in presence of 0.10 g·L-1 polyethylene glycol and 40 mL·L-1 formic acid mixture as surfactant. A very smooth, shiny and crack free Cr coating has been obtained at applied deposition potentials up to -0.70 V and current densities up to -15 A·m-2 between 50 0C and 80 0C. The cathodic current efficiency for the deposition of chromium is about 98.4%. Key words: Electrodeposition; Reline; Ethaline; Chromium; Cyclic voltammetry

EM-134 Application and Future Prospects of CRISPR-Case in Food and Agriculture Science to Meet Zero Hunger Goal Paroma Arefin 1*, Aishawarya Arefin2, Md Saidul Arefin 3 1 Scientific Officer, BCSIR Laboratories, Rajshahi 2 Institute of Health Economics, University of Dhaka 3 Professor, Institute of Nutrition and Food Science, University of Dhaka Corresponding author email: [email protected]

The CRISPR-Cas9 tool is a selective genome editing technology developed from bacterial immune system of type II and can be spread to almost all species. The invention of the CRISPR-Cas9 methods are promising due to their relative efficiency, ease of use, comparatively low cost, diversified editing prospects in a single procedure. CRISPR-Cas9 can be utilized for the improvement of agronomic traits including enhanced crop yield, disease tolerance, regulation of gene expression, combat biotic and abiotic stresses, and generation of genome-wide mutant libraries. Advancement of this technology in food and agriculture sector is still almost in primary stage. But CRISPR-Cas9 based genome editing is going to be the key technique for proper plant editing and thus helping to achieve the zero hunger goal of SDG and arranging food for the growing human population. In this study, we will discuss current progresses in the utilization of the CRISPR-Cas9 system in the food and agricultural industry, particularly in the development of resistant crops with improved quality and productivity. Additionally we have explored the state of the global food market and discuss the safety and ethical concerns associated with the application of this technology in the food and agricultural industry.

EM-135 Investigation of fatty acids composition of four different seed`s oil (Peanut, Neem, Coconut and Nigella) grown in Bangladesh and its pharmacological and physiological role in human body Md. Abdurrahim1,Md. Asadul Islam1*, Md. Abdul Jalil1, Md. Badrul Islam1, Mst. Sarmina Yeasmin2,Nazim Uddin Ahmed1, Md. Mahmudul Hassan Mondol1, Ali Ahsan Muzahid1, Atoll Goswami1, Nafisa Akter2andMd. Abdul Jalil2 Drugs and Toxins Research Division1, BCSIR Laboratories, Rajshahi. Oils, Fats and Waxes Research Division2, BCSIR Laboratories, Rajshahi.

Fats and oils are the main reserve energy source in plethora of animals, especially in mammals. In this work we investigated fatty acid composition of four seed oils in terms of eight fatty acids. Among them, Nigela oil has been found to have highest percentage (49.57) of linoleic acid (ω-6) followed by PO (47.89), NeO (30.46) and CO (10.07). Oleic acid (ω-9) was found to be highest in PO (34.96%) followed by 27, 26 and 12% in PO, NiO and CO. The variation in amount of eight different fatty acids in coconut oil is relatively lower than that of others. Eight different types of fatty acids were analyzed quantitatively for each of the oils. Six of them were saturated, one was ω-6 which is a member of PUFA and another was one ω-9 MUFA. Study shows the fatty acid composition of four different fixed oils. Among them, Nigela oil (NiO) has been found to have highest percentage (49.57) of linoleic acid followed by Peanut Oil (PO) (47.89), Neem Oil (NeO) (30.46) and Coconut Oil (CO) (10.07). Oleic acid (a ω-9 unsaturated fatty acid) was found to be highest in PO (34.96%) followed by 27, 26 and 12% in PO, NiO and CO. The variation in amount of eight different fatty acids in coconut oil is relatively lower than that of others which indicates that all the responsible genes for synthesis of the fatty acids under investigation are well expressed in coconut. Keywords: Fatty Acid composition, PUFA, MUFA, Oil seeds, GC-MS

212

EM-136 A Statistical Investigation on the Risk Factors of Cancer Md. Masud Rana*, Md. Hadiul Kabir and M. Nurul Haque Mollah Bioinformatics Lab., Dept. of Statistics, Rajshahi University, Bangladesh *Poster presenting author ([email protected])

Cancer is a group of diseases involving abnormal cell growth with the potential to invade or spread to other parts of the body. It is the second leading cause of death in this world and accounted for 7.4 million deaths (around 13% of all deaths) according to the WHO reports 2009. The different types of cancer are lung (1.3 million deaths/year), colorectal (639,000 deaths/year), stomach (803,000 deaths/year), breast (519,000 deaths), liver (610,000 deaths/year), and so on. More than 70% of all cancer deaths occurred in low and middle income countries. According to WHO reports, the significant risk factors are smoking, alcohol drinking, fatty food intake, obesity, physical inactivity, chronic infections diseases, environmental and occupational risks including ionizing and non-ionizing radiation. However, some environmental issues reduce the influence of the risk factor of cancer diseases. In this work, an attempt is made to validate those risk factors of cancer diseases in our local environment. Our real data analysis by contingency table and logistic regression model showed that genetic heredity, smoking, fatty food intake, obesity and physical inactivity are the significant risk factor of cancer diseases. We hope the output of this research project will contribute to the awareness of our country peoples about cancer. Keywords: Cancer, Risk factor, Contingency analysis, Logistic regression analysis, Significance test and awareness.

213