DOCSLIB.ORG

Welcome to F-Secure Capital Markets

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Welcome to F-Secure Capital Markets TECHNOLOGY F-SECURE’S UNIQUE CAPABILITIES IN DETECTION & RESPONSE Jyrki Tulokas, EVP, Cyber security products & services UNDERSTANDING THE THREAT LANDSCAPE Human orchestration Nation states NATION STATE Unknown 0-day tools and ATTACKS techniques Organized cyber crime TARGETED Human conducted, stealthy, ATTACKS targeted attacks ADVANCED Cyber crime THREATS Fileless scripts, System tools, Ransomware MASS Commodity threats ATTACKS Opportunistic phishing Malware Spam Technology drives the scale Volume 2 CUSTOMERS CONTINUE TO NEED BOTH PREVENTIVE AND REACTIVE CAPABILITIES PREVENTIVE DETECTION & RESPONSE RECON WEAPONIZE DELIVER EXPLOIT CONTROL EXECUTE MAINTAIN TACTICS, TECHNIQUES AND PROCEDURES Persistence Discovery Exfiltration Privilege Escalation Lateral Movement Command Defense Evasion Execution and Control Credential Access Collection 3 Targeted Attack Simulation Incident response Threat Hunting INDUSTRY LEADING RESEARCH AND THREAT VISIBILITY AUTONOMOUS SELF LEARNING F-Secure Labs BIG DATA PLATFORM Forensics Corrective actions Endpoint data Big Data and AI platform Actionable is essential insights and response 4 AUTONOMOUS BIG DATA DETECTION & RESPONSE PLATFORM BUILT FOR MASSIVE SCALE DATA COLLECTION SENSORS REAL-TIME ANALYTICS IN THE CLOUD The endpoint sensors collect : Cloud used for analytics: ✓ file accesses; ✓ Machine learning ✓ process creations; ✓ Broad Context Detection ™ ✓ network connections; ✓ Automatic analysis, categorization and ✓ registry writes; detection creation ✓ system log entries relevant to ✓ Telemetry of the attacks shared in real- detecting security breaches; time with our security cloud ✓ extracts of scripts derived from ✓ Long-term view to threat propagation run-time execution; EVENTS/MONTH66 BN 5 ANOVA, Active learning, AdaBoost, Analogical modeling, Anomaly detection, Apriori algorithm, Artificial neural network, Association rule learning algorithms, Averaged One-Dependence Estimators (AODE), BIRCH, Back-Propagation, Bayesian Belief Network (BBN), Bayesian Network (BN), Bayesian knowledge base, Bias-variance dilemma, Binary classifier, Boosting, Bootstrap aggregating (Bagging), C4.5 algorithm, C5.0 algorithm, Canonical correlation analysis (CCA), CaseA -SELFbased reasoning,-LEARNINGChi-squared AutomaticAIInteraction PLATFORMDetection (CHAID), Classification and regression tree (CART), Co-training, Conceptual clustering, Conditional Random Field, Conditional decision tree, Conditional random field (CRF), Convolutional Neural Network (CNN), DBSCAN, ALREADYData Pre-processing, DecisionPROVENstump, Decision ONtree, THEDeep Belief MARKETNetworks (DBN), Deep Boltzmann Machine (DBM), Deep Convolutional neural networks, Deep Recurrent neural networks, Eclat algorithm, Elastic net, Empirical risk minimization, Ensemble averaging, Ensembles of classifiers, Expectation Maximization (EM), Extreme learning machine, FP-growth algorithm, Factor analysis, Feature engineering, Feature extraction, Feature learning, Feature selection, Feedforward neural network, Fisher's linear discriminant, Flexible Discriminant Analysis (FDA), Fuzzy clustering, Gaussian Naive Bayes, Gaussian process regression, Gene expression programming, Generative models, Generative Adversial Network (GAN), Generative topographic map, Gradient boosted decision tree (GBDT), Gradient boosting machine (GBM), Graph-based methods, Graphical models,SUPERVISEDGroup method of MLdata handling (GMDH),UNSUPERVISEDHidden Markov model (HM), ML Hierarchical Clustering, HierarchicalDATA classifier, Hierarchical hidden Markov model,MODULESHierarchical temporal memory, Hopfield Network,MODULESID3, Independent component analysisTRANSFORMATION(ICA), Inductive bias, Inductive logic programming, Information bottleneck method, Information fuzzy networks (IFN), Instance-based learning, Iterative Dichotomiser 3 (ID3), K-means clustering, K-•mediansDetectingclustering,threatsK-nearest neighbors algorithm• User/host(KNN), Lazy profilinglearning, Learning Automata,• LearningAdvancedVector Quantization (LVQ), Learning to rank, Least Absolute Shrinkage and Selection•Operator (LASSO), Least-angle regression (LARS), Linear classifier, Linear discriminant analysis (LDA), Linearbasedregression, on trainingLocal outlierdata factor, Logic learningAnomalymachine, detectionLogistic Model Tree, Logistic regression,visualizationLocally Estimated Scatterplot Smoothing (LOESS),• LearningLocally Weightedfrom expertLearning (LWL), Long Short-Term Memory Recurrent Neural Networks• Intelligently(LSTM), Low-density clustersseparation, M5, Mean-shift, Metadata,feedbackMinimum message length, Mixture Discriminant Analysis (MDA), Multi-label classification,anomaliesMultidimensional scaling (MDS), Multinomial Naive Bayes, Multinomial logistic regression, Multivariate adaptive regression splines (MARS), Naive Bayes, Nearest Neighbor Algorithm, Non-negative matrix factorization (NMF), OPTICS algorithm, Occam learning, Online machine learning, Ordinal classification, Ordinary least squares regression (OLSR), PAC learning, Partial least squares regression (PLSR), Perceptron, Principal component analysis (PCA), Principal component regression (PCR), Probabilistic classifier, Probably approximately correct learning (PAC) learning, Projection pursuit, Q-learning, Quadratic classifiers, Quadratic Discriminant Analysis (QDA), Random Forest, Radial Basis Function Network (RBFN), Regression, Reinforcement Learning, Ridge regression, Ripple down rules, a knowledge acquisition methodology, SLIQ, SPRINT, Sammon mapping, Self-organizing map (SOM), Semi-supervised learning, Single-linkage clustering, Stacked Auto-Encoders, Stacked Generalization (blending), State–action–reward–state–action (SARSA), Statistical learning, Stepwise regression, Structured prediction, Support vector machine (SVM), Symbolic machine learning algorithms, t-distributed stochastic neighbor embedding (t-SNE), Temporal difference learning (TD), Transduction, Unsupervised learning, VC theory, Vector Quantization 6 RESULT: INDUSTRY’S FASTEST AND MOST ACCURATE DETECTIONS 900,000 SUSPICIOUS EVENTS 25 15 DETECTIONS REAL THREATS 2 billion • Event enrichment DATA EVENTS/MONTH • Detections of Customer confirmed Host & User Profiling which customer that these were • Anomaly Detection was notified. real threats • Detection Significance Analysis 7 F-SECURE COUNTERCEPT: HUMAN AUGMENTED TECHNOLOGY, DEEP INSIGHT TO TARGETED ATTACKS The hunt team use ‘Explore’ and ‘Investigate/Respond’. They create new ‘assisted hunts’ and tags which propagate to all other instances of THP including those used by our clients and our own 24/7 detection and response team. AVAILABLE IN THREE MODELS: • As a service • Assisted • Customer specific 8 8 TOWARD SELF-HEALING SYSTEMS AND GUIDED RESPONSE • Recommends response actions of RECOMMENDED F-SECURE THREAT RESPONSE ACTIONS HUNTER’S GUIDANCE informing users, or isolating hosts High risk Acknowledged • Get help on tough investigations Medium Confidence Jan 12, 2018 12:34:56 High Criticality from F-Secure experts with Elevate to F-Secure Inform users Inform admins 3 similar Recent BC detections • Constantly improves Isolate hosts recommendations and detections with machine learning Recommended actions Elevate to F-Secure 9 SUMMARY • Aiming to be a key player in detection and response technology and services • Best-in-class integrated cyber security suite • Scalability through managed services partners • Security for the cloudified world 10.
Load more

Recommended publications
  • Machine Learning Or Econometrics for Credit Scoring: Let’S Get the Best of Both Worlds Elena Dumitrescu, Sullivan Hué, Christophe Hurlin, Sessi Tokpavi
    Machine Learning or Econometrics for Credit Scoring: Let’s Get the Best of Both Worlds Elena Dumitrescu, Sullivan Hué, Christophe Hurlin, Sessi Tokpavi To cite this version: Elena Dumitrescu, Sullivan Hué, Christophe Hurlin, Sessi Tokpavi. Machine Learning or Econometrics for Credit Scoring: Let’s Get the Best of Both Worlds. 2020. hal-02507499v2 HAL Id: hal-02507499 https://hal.archives-ouvertes.fr/hal-02507499v2 Preprint submitted on 2 Nov 2020 (v2), last revised 15 Jan 2021 (v3) HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Machine Learning or Econometrics for Credit Scoring: Let's Get the Best of Both Worlds∗ Elena, Dumitrescu,y Sullivan, Hu´e,z Christophe, Hurlin,x Sessi, Tokpavi{ October 18, 2020 Abstract In the context of credit scoring, ensemble methods based on decision trees, such as the random forest method, provide better classification performance than standard logistic regression models. However, logistic regression remains the benchmark in the credit risk industry mainly because the lack of interpretability of ensemble methods is incompatible with the requirements of financial regulators. In this paper, we pro- pose to obtain the best of both worlds by introducing a high-performance and inter- pretable credit scoring method called penalised logistic tree regression (PLTR), which uses information from decision trees to improve the performance of logistic regression.
    [Show full text]
  • Full Academic Cv: Grigori Fursin, Phd
    FULL ACADEMIC CV: GRIGORI FURSIN, PHD Current position VP of MLOps at OctoML.ai (USA) Languages English (British citizen); French (spoken, intermediate); Russian (native) Address Paris region, France Education PhD in computer science with the ORS award from the University of Edinburgh (2004) Website cKnowledge.io/@gfursin LinkedIn linkedin.com/in/grigorifursin Publications scholar.google.com/citations?user=IwcnpkwAAAAJ (H‐index: 25) Personal e‐mail [email protected] I am a computer scientist, engineer, educator and business executive with an interdisciplinary background in computer engineering, machine learning, physics and electronics. I am passionate about designing efficient systems in terms of speed, energy, accuracy and various costs, bringing deep tech to the real world, teaching, enabling reproducible research and sup‐ porting open science. MY ACADEMIC RESEARCH (TENURED RESEARCH SCIENTIST AT INRIA WITH PHD IN CS FROM THE UNIVERSITY OF EDINBURGH) • I was among the first researchers to combine machine learning, autotuning and knowledge sharing to automate and accelerate the development of efficient software and hardware by several orders of magnitudeGoogle ( scholar); • developed open‐source tools and started educational initiatives (ACM, Raspberry Pi foundation) to bring this research to the real world (see use cases); • prepared and tought M.S. course at Paris‐Saclay University on using ML to co‐design efficient software and hardare (self‐optimizing computing systems); • gave 100+ invited research talks; • honored to receive the
    [Show full text]
  • Reinforcement Learning in Supervised Problem Domains
    Technische Universität München Fakultät für Informatik Lehrstuhl VI – Echtzeitsysteme und Robotik reinforcement learning in supervised problem domains Thomas F. Rückstieß Vollständiger Abdruck der von der Fakultät für Informatik der Technischen Universität München zur Erlangung des akademischen Grades eines Doktors der Naturwissenschaften (Dr. rer. nat.) genehmigten Dissertation. Vorsitzender: Univ.-Prof. Dr. Daniel Cremers Prüfer der Dissertation 1. Univ.-Prof. Dr. Patrick van der Smagt 2. Univ.-Prof. Dr. Hans Jürgen Schmidhuber Die Dissertation wurde am 30. 06. 2015 bei der Technischen Universität München eingereicht und durch die Fakultät für Informatik am 18. 09. 2015 angenommen. Thomas Rückstieß: Reinforcement Learning in Supervised Problem Domains © 2015 email: [email protected] ABSTRACT Despite continuous advances in computing technology, today’s brute for- ce data processing approaches may not provide the necessary advantage to win the race against the ever-growing amount of data that can be wit- nessed over the last decades. In this thesis, we discuss novel methods and algorithms that are capable of directing attention to relevant details and analysing it in sequence to overcome the processing bottleneck and to keep up with this data explosion. In the first of three parts, a novel exploration technique for Policy Gradi- ent Reinforcement Learning is presented which replaces traditional ad- ditive random exploration with state-dependent exploration, exploring on a higher, more strategic level. We will show how this new exploration method converges faster and finds better global solutions than random exploration can. The second part of this thesis will introduce the concept of “data con- sumption” and discuss means to minimise it in supervised learning tasks by deriving classification as a sequential decision process and ma- king it accessible to Reinforcement Learning methods.
    [Show full text]
  • Delivering a Machine Learning Course on HPC Resources
    Delivering a machine learning course on HPC resources Stefano Bagnasco, Federica Legger, Sara Vallero This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement LHCBIGDATA No 799062 The course ● Title: Big Data Science and Machine Learning ● Graduate Program in Physics at University of Torino ● Academic year 2018-2019: ○ Starts in 2 weeks ○ 2 CFU, 10 hours (theory+hands-on) ○ 7 registered students ● Academic year 2019-2020: ○ March 2020 ○ 4 CFU, 16 hours (theory+hands-on) ○ Already 2 registered students 2 The Program ● Introduction to big data science ○ The big data pipeline: state-of-the-art tools and technologies ● ML and DL methods: ○ supervised and unsupervised models, ○ neural networks ● Introduction to computer architecture and parallel computing patterns ○ Initiation to OpenMP and MPI (2019-2020) ● Parallelisation of ML algorithms on distributed resources ○ ML applications on distributed architectures ○ Beyond CPUs: GPUs, FPGAs (2019-2020) 3 The aim ● Applied ML course: ○ Many courses on advanced statistical methods available elsewhere ○ Focus on hands-on sessions ● Students will ○ Familiarise with: ■ ML methods and libraries ■ Analysis tools ■ Collaborative models ■ Container and cloud technologies ○ Learn how to ■ Optimise ML models ■ Tune distributed training ■ Work with available resources 4 Hands-on ● Python with Jupyter notebooks ● Prerequisites: some familiarity with numpy and pandas ● ML libraries ○ Day 2: MLlib ■ Gradient Boosting Trees GBT ■ Multilayer Perceptron Classifier MCP ○ Day 3: Keras ■ Sequential model ○ Day 4: bigDL ■ Sequential model ● Coming: ○ CUDA ○ MPI ○ OpenMP 5 ML Input Dataset for hands on ● Open HEP dataset @UCI, 7GB (.csv) ● Signal (heavy Higgs) + background ● 10M MC events (balanced, 50%:50%) ○ 21 low level features ■ pt’s, angles, MET, b-tag, … Signal ○ 7 high level features ■ Invariant masses (m(jj), m(jjj), …) Background: ttbar Baldi, Sadowski, and Whiteson.
    [Show full text]
  • Analysis of Earthquake Forecasting in India Using Supervised Machine Learning Classifiers
    sustainability Article Analysis of Earthquake Forecasting in India Using Supervised Machine Learning Classifiers Papiya Debnath 1, Pankaj Chittora 2 , Tulika Chakrabarti 3, Prasun Chakrabarti 2,4, Zbigniew Leonowicz 5 , Michal Jasinski 5,* , Radomir Gono 6 and Elzbieta˙ Jasi ´nska 7 1 Department of Basic Science and Humanities, Techno International New Town Rajarhat, Kolkata 700156, India; [email protected] 2 Department of Computer Science and Engineering, Techno India NJR Institute of Technology, Udaipur 313003, Rajasthan, India; [email protected] (P.C.); [email protected] (P.C.) 3 Department of Basic Science, Sir Padampat Singhania University, Udaipur 313601, Rajasthan, India; [email protected] 4 Data Analytics and Artificial Intelligence Laboratory, Engineering-Technology School, Thu Dau Mot University, Thu Dau Mot City 820000, Vietnam 5 Department of Electrical Engineering Fundamentals, Faculty of Electrical Engineering, Wroclaw University of Science and Technology, 50-370 Wroclaw, Poland; [email protected] 6 Department of Electrical Power Engineering, Faculty of Electrical Engineering and Computer Science, VSB—Technical University of Ostrava, 708 00 Ostrava, Czech Republic; [email protected] 7 Faculty of Law, Administration and Economics, University of Wroclaw, 50-145 Wroclaw, Poland; [email protected] * Correspondence: [email protected]; Tel.: +48-713-202-022 Abstract: Earthquakes are one of the most overwhelming types of natural hazards. As a result, successfully handling the situation they create is crucial. Due to earthquakes, many lives can be lost, alongside devastating impacts to the economy. The ability to forecast earthquakes is one of the biggest issues in geoscience. Machine learning technology can play a vital role in the field of Citation: Debnath, P.; Chittora, P.; geoscience for forecasting earthquakes.
    [Show full text]
  • B.Sc Computer Science with Specialization in Artificial Intelligence & Machine Learning
    B.Sc Computer Science with Specialization in Artificial Intelligence & Machine Learning Curriculum & Syllabus (Based on Choice Based Credit System) Effective from the Academic year 2020-2021 PROGRAMME EDUCATIONAL OBJECTIVES (PEO) PEO 1 : Graduates will have solid basics in Mathematics, Programming, Machine Learning, Artificial Intelligence fundamentals and advancements to solve technical problems. PEO 2 : Graduates will have the capability to apply their knowledge and skills acquired to solve the issues in real world Artificial Intelligence and Machine learning areas and to develop feasible and reliable systems. PEO 3 : Graduates will have the potential to participate in life-long learning through the successful completion of advanced degrees, continuing education, certifications and/or other professional developments. PEO 4 : Graduates will have the ability to apply the gained knowledge to improve the society ensuring ethical and moral values. PEO 5 : Graduates will have exposure to emerging cutting edge technologies and excellent training in the field of Artificial Intelligence & Machine learning PROGRAMME OUTCOMES (PO) PO 1 : Develop knowledge in the field of AI & ML courses necessary to qualify for the degree. PO 2 : Acquire a rich basket of value added courses and soft skill courses instilling self-confidence and moral values. PO 3 : Develop problem solving, decision making and communication skills. PO 4 : Demonstrate social responsibility through Ethics and values and Environmental Studies related activities in the campus and in the society. PO 5 : Strengthen the critical thinking skills and develop professionalism with the state of art ICT facilities. PO 6 : Quality for higher education, government services, industry needs and start up units through continuous practice of preparatory examinations.
    [Show full text]
  • Automatic Classification of Oranges Using Image Processing and Data Mining Techniques
    Automatic classification of oranges using image processing and data mining techniques Juan Pablo Mercol, Juliana Gambini, Juan Miguel Santos Departamento de Computaci´on, Facultad de Ciencias Exactas y Naturales, UBA, Buenos Aires, Argentina [email protected],{jgambini,jmsantos}@dc.uba.ar, Abstract Data mining is the discovery of patterns and regularities from large amounts of data using machine learning algorithms. This can be applied to object recognition using image processing techniques. In fruits and vegetables production lines, the quality assurance is done by trained people who inspect the fruits while they move in a conveyor belt, and classify them in several categories based on visual features. In this paper we present an automatic orange’s classification system, which uses visual inspection to extract features from images captured with a digital camera. With these features train several data mining algorithms which should classify the fruits in one of the three pre-established categories. The data mining algorithms used are five different decision trees (J48, Classification and Regression Tree (CART), Best First Tree, Logistic Model Tree (LMT) and Random For- est), three artificial neural networks (Multilayer Perceptron with Backpropagation, Radial Basis Function Network (RBF Network), Sequential Minimal Optimization for Support Vector Machine (SMO)) and a classification rule (1Rule). The obtained results are encouraging because of the good accuracy achieved by the clas- sifiers and the low computational costs. Keywords: Image processing, Data Mining, Neural Networks, Decision trees, Fruit Qual- ity Assurance, Visual inspection, Artificial Vision. 1 Introduction During the last years, there has been an increase in the need to measure the quality of several products, in order to satisfy customers needs in the industry and services levels.
    [Show full text]
  • A Gradient-Based Split Criterion for Highly Accurate and Transparent Model Trees
    A Gradient-Based Split Criterion for Highly Accurate and Transparent Model Trees Klaus Broelemann and Gjergji Kasneci SCHUFA Holding AG, Wiesbaden, Germany fKlaus.Broelemann,[email protected] Abstract As a consequence, the choice of a predictive model for a specific task includes a trade-off between simplicity and Machine learning algorithms aim at minimizing the transparency on one side and complexity and higher predic- number of false decisions and increasing the accu- tive power on the other side. If the task is mainly driven by racy of predictions. However, the high predictive the quality of the prediction, complex state-of-the-art models power of advanced algorithms comes at the costs are best suited. Yet, many applications require a certain level of transparency. State-of-the-art methods, such as of transparency. This can be driven by regulatory constraints, neural networks and ensemble methods, result in but also by the user’s wish for transparency. A recent study highly complex models with little transparency. shows that model developers often feel the pressure to use We propose shallow model trees as a way to com- transparent models [Veale et al., 2018], even if more accu- bine simple and highly transparent predictive mod- rate models are available. This leads to simple models, such els for higher predictive power without losing the as linear or logistic regression, where more complex models transparency of the original models. We present would have much higher predictive power. In other cases, a novel split criterion for model trees that allows transparency is not required, but optional.
    [Show full text]
  • Delivering a Machine Learning Course on HPC Resources
    EPJ Web of Conferences 245, 08016 (2020) https://doi.org/10.1051/epjconf/202024508016 CHEP 2019 Delivering a machine learning course on HPC resources Stefano Bagnasco1, Gabriele Gaetano Fronzé1, Federica Legger1;∗ Stefano Lusso1, and Sara Vallero1 1Istituto Nazionale di Fisica Nucleare, via Pietro Giuria 1, 10125 Torino, Italy Abstract. In recent years, proficiency in data science and machine learning (ML) became one of the most requested skills for jobs in both industry and academy. Machine learning algorithms typically require large sets of data to train the models and extensive usage of computing resources, both for training and inference. Especially for deep learning algorithms, training performances can be dramatically improved by exploiting Graphical Processing Units (GPUs). The needed skill set for a data scientist is therefore extremely broad, and ranges from knowledge of ML models to distributed programming on heterogeneous resources. While most of the available training resources focus on ML algorithms and tools such as TensorFlow, we designed a course for doctoral students where model training is tightly coupled with underlying technologies that can be used to dynamically provision resources. Throughout the course, students have access to a dedicated cluster of computing nodes on local premises. A set of libraries and helper functions is provided to execute a parallelized ML task by automatically deploying a Spark driver and several Spark execution nodes as Docker containers. Task scheduling is managed by an orchestration layer (Kubernetes). This solution automates the delivery of the software stack required by a typical ML workflow and enables scalability by allowing the execution of ML tasks, including training, over commodity (i.e.
    [Show full text]
  • Iowa State University – 2016-2017 1
    Iowa State University – 2016-2017 1 CPR E 281 Digital Logic 4 COMPUTER SCIENCE COM S 309 Software Development Practices 3 COM S 311 Design and Analysis of Algorithms 3 http://www.cs.iastate.edu COM S 321 Introduction to Computer Architecture and 3 The undergraduate curriculum in computer science leading to the degree Machine-Level Programming Bachelor of Science is accredited by the Computing Accreditation COM S 327 Advanced Programming Techniques 3 Commission of ABET, http://www.abet.org , and equips students with COM S 331 Theory of Computing 3 a sound knowledge of the foundations of computer science, as well COM S 342 Principles of Programming Languages 3 as the problem solving and system design skills necessary to create robust, efficient, reliable, scalable, and flexible software systems. The COM S 352 Introduction to Operating Systems 3 B.S. degree in Computer Science prepares students for graduate study COM S 362 Object-Oriented Analysis and Design 3 in computer science, and for various business, industry, and government positions including computer scientists, information technologists, In addition to the above courses, at least 6 credits of 400-level courses and software developers. The main educational objectives of the are required, with a grade of C- or better. At least 3 credits must be from computer science program at Iowa State University are that its graduates courses in Group 1 (oral and written reports) and the remaining credits demonstrate expertise, engagement, and learning within three to five from courses in Group 1 or 2. Com S 402 may be applied towards the years after graduation.
    [Show full text]
  • Accurate and Intuitive Contextual Explanations Using Linear Model
    Accurate and Intuitive Contextual Explanations using Linear Model Trees Aditya Lahiri Narayanan Unny Edakunni American Express, AI Labs American Express, AI Labs Bangalore, Karnataka, India Bangalore, Karnataka, India [email protected] [email protected] Abstract 1 Introduction With the ever-increasing use of complex machine learning The Finance domain has a number of scenarios where sophis- models in critical applications within the finance domain, ticated machine learning models are used to predict various explaining the modelâĂŹs decisions has become a neces- quantities of interest. This ranges from prediction of credit sity. With applications spanning from credit scoring[18] to and fraud risk of transactions to digital marketing. In many credit marketing[1], the impact of these models is undeni- of these instances, it is not enough to provide accurate pre- able. Among the multiple ways in which one can explain dictions, but is also important to provide an intuitive reason the decisions of these complicated models[16], local post in coming up with the prediction that can be understood by hoc model agnostic explanations have gained massive adop- a human. In instances like credit approval, it is also man- tion. These methods allow one to explain each prediction dated by law to provide an explanation to consumer for the independent of the modelling technique that was used while decision made for approval or denial of the credit[8, 13]. In training. As explanations, they either give individual feature applications like credit line extension[14], predictions from attributions[20] or provide sufficient rules that represent these sophisticated machine learnt models could be used by conditions for a prediction to be made[21].
    [Show full text]
  • A Classification Model for Water Quality Analysis
    European Journal of Computer Science and Information Technology Vol.7, No.3, pp.1-8, June 2019 ___Published by European Centre for Research Training and Development UK (www.eajournals.org) A CLASSIFICATION MODEL FOR WATER QUALITY ANALYSIS USING DECISION TREE Consolata Gakii and Jeniffer Jepkoech Department of Mathematics, Computer science and Information Technology (MCIT). University of Embu P.O Box 06 Embu ABSTRACT: A classification algorithm is used to assign predefined classes to test instances for evaluation) or future instances to an application). This study presents a Classification model using decision tree for the purpose of analyzing water quality data from different counties in Kenya. The water quality is very important in ensuring citizens get to drink clean water. Application of decision tree as a data mining method to predict clean water based on the water quality parameters can ease the work of the laboratory technologist by predicting which water samples should proceed to the next step of analysis. The secondary data from Kenya Water institute was used for creation of this model. The data model was implemented in WEKA software. Classification using decision tree was applied to classify /predict the clean and not clean water. The analysis of water Alkalinity,pH level and conductivity can play a major role in assessing water quality. Five decision tree classifiers which are J48, LMT, Random forest, Hoeffding tree and Decision Stump were used to build the model and the accuracy compared. J48 decision tree had the highest accuracy of 94% with Decision Stump having the lowest accuracy of 83%. KEYWORDS: Data Mining, classification model, Decision tree, Weka Tool, water quality INTRODUCTION Supervised learning is a machine learning algorithm which receives feature vector and the target pattern as an input to build a model.
    [Show full text]