Business Analytics, ML, AI - Course Topics

This course will be done in 2 parts – Business Analytics and ML/AI.

Each course has 130 hours + 130 hours of classroom lecture and practical work. (total 260 hours)

Business Analytics course will be conducted over 4.5 months, and those who qualified in the basic course will be eligible to join the advanced course. ML/AI course will be conducted over 4.5 months. Thus, the total duration for both will be 9 calendar months.

Multiple trainers will take classes.

Each course will have 1 ungraded assignment, 1 graded assignment and 1 exam.

Audience must know at least 1 programming language (basic level is sufficient) and 1 (basic level is sufficient).

Minimum attendance of 80% is a must.

There will be 2 mock selection tests and 1 mock interview to candidates who score well in the graded assignments and exams. A set of HR members of different companies will be contacted to evaluate candidates for any openings in their organization, as part of a job mela. (This is optional for candidates)

Foundation Session – 1 week

Before students join the course, they have to go thru a warm-up or foundation session. The foundation session will be conducted for 1 week. The aim of this is to set the expectations and the necessary spade works properly. Students will get a clarity on what they will get out of this, the opportunities, the basics that they need to refresh etc.

1. The role of data in decision making 2. Day-to-day data generation and data consumption 3. Data manipulation with Excel 4. Evolution of data analytics space 5. Market needs and opportunities 6. What are the current skillset gaps? 7. Our expectations from you 8. Basic infrastructure requirements 9. How to make the best use of online education? 10. How to develop passion towards data?

Course: Business Analytics – 4.5 months – 130 hours

Basics of Data Science (10 hours) • What are business intelligence and business analytics? • Need for data science • Introduction to bigdata • Data freedom • How to retrieve data and process data • Role of data scientist vs data analyst • What you do not obviously see in data, unearthing the hidden facts • Concepts of predictive and prescriptive analytics

Data Management Fundamentals using SQL (25 hours) • What are OLTP and OLAP? • Master data, transaction data • Calculating the volume of transactions in terms of number of records and storage size • Ingestion of data to data lake (structured data from subsystems, ETL - Extract Transform Load process, initial data load and incremental data load) • Abstraction of data, use of meta data • Typical system and hardware architecture of analytics solution for enterprise • Moving to Big Data and NoSQL data bases • Postgres open source database o Create database, tables, indexes (to ensure non-development people are brought to the same level with development teams) o Select, insert, update, delete statements, conditions, groupings, sorting o Joins, aggregate functions, load from file, dump to file • Basic usage of new generation bigdata ClickHouse, a superfast database

Business Insights (10 hours) • Need for business metrics (for different roles /levels of people in enterprise) • Real time metrics, near real time metrics, offline metrics • Decision enabling and decision making based on data - examples • The first 100 business metrics for any specific industry • EDA - Exploratory data analysis • Aggregate metrics - count of transactions, sum of transactions, distribution of transactions based on types

Data Visualization (25 hours) • Getting data to data mart for faster visuals (connect to metabase and sync) • Metabase open source o Data types, various chart types, Automated dashboards, custom queries o Plotting data based on classifications/ buckets, time, trending o Drill downs to provide more details, filters o Security features – users, roles, access o Pulse • Tableau o Basic visuals, filters, groups, Aggregates o Drill downs, dashboards o Maps, sheets

Domain Specific Business Analytics (20 hours) • Data Samples: Core banking, ATMs, Net banking, loans, cards etc., point of sale data, TV viewership data, News feeds text, sports data in real time etc. • Common metrics: Customer-based metrics, employee-based metrics, brand/service-based metrics, geography-based metrics • Banking - Patterns of new account openings, account closures, payments and defaulting • eCom- Customer profiling - age buckets, profession bins etc., purchase profile • CRM - Campaign based metrics on products, measuring campaign success, measuring new sell vs up sell success • HRMS - Employee training metrics, feedback metrics • Retail - Point of sale analytics • Government - Smart city analytics • Anonymize data by 2 level translations and encryption • eCom - Customer feedback analytics • Trend analysis - transactions, customers, products etc. • of Things (IoT)/Sensors metrics – Air pollution data, plant sensor data • Telecom - Call Detail Records (CDR)

Python programming (20 hours) • Comments, variables Operations – numeric, string, logical • Conditions, loops, Lists, Tuples, dictionaries, sets • File handling – open, close, read, write • Date operations, exception handling • Database operations – fetch data for analytics only and not update/delete/insert • Functions, parameters • Basic numpy operations – like avg, mean median mode, std dev, plot, random • Matplotlib library • Linear algebraic functions. Matrices, Determinants, Inverses • Logic building tips • Plotting – pie, bar, scatter, line, multi charts

Basic Statistics using Python (15 hours) • Preparing IDE and loading data • Examples on Exploratory analysis - numerical summary, rule, plots, multi variate • Examples on standard mathematical analysis - mean/median/percentile/variance, distribution based on frequency/poisson/sampling • Examples on statistics analysis – descriptive, inferential, correlation, error types • Running regular statistics packages for average, 98th percentile, etc.

Project Management (5 hours) • Checklist for Analytics project • Gaining Subject Matter Expertise • Key Process Indicators for different industries • Typical task list for an analytics project

Project work (30 hours – outside class hours) • Ingest large sets of data in postgres or ClickHouse • Carry out exploratory analytics • Build aggregates and data marts • Build visuals and dashboards using tableau or Metabase • Present the business metrics

Course: Machine Learning & Artificial Intelligence – 4.5 months – 130 hours

Introduction to Machine Learning (10 hours) • What cannot be done by manual analysis? • Concepts of Supervised learning • Training data, test data - banking customer footprints, ATM and net access data • Applying Clustering, representation learning • Unsupervised learning - Clustering, Neural network, convolutional neural network • Introduction to artificial neural networks

Mathematics for Machine Learning (15 hours) • Linear Algebra • Probability Theory, Eigenvalues and Eigenvectors • Principal Component Analysis • Miscellaneous mathematical topics, line of best fit

Advanced Statistics using Python (25 hours) • Intro to statistics • Identifying the dependent data and independent data • Running linear regression on financial data sets • Population parameter estimation • Confidence interval estimation • Hypothesis testing • Ttest and ztest

Regression & Classification for Business Applications (50 hours) Time series forecasting • Exponential Smoothening • ARIMA, Auto ARIMA

Regression • Multiple Linear Regression • Subset Regression • Regularization - bias, variances, lasso, ridge and elastic net

Classification • Logistic Regression • SVM • Naïve Bayes classification • Random Forest • KNN

Clustering • K Means Clustering • Fuzzy K Means Clustering

AutoML • H20

Advanced Topics (TensorFlow) (30 Hours) • Setting up TensorFlow virtual environment for python • Creating Neural Network with Keras • Constructing Models in Keras • Employing Layers in Keras Models • Building Convolutional NN with Keras • Introduction to deep learning • Regression and prediction using ANN • Image Classification using CNN

Project work (30 hours – outside class hours) • Ingest large sets of data in postgres or ClickHouse • Carry out exploratory analytics • Build visuals and dashboards using Tableau or Metabase • Create prediction model • Present results using multiple models