Business Analytics

A course focused on building quantitative models and applying statistical techniques to solve complex business problems and make better decisions.

This course was designed to build the critical skill of creating quantitative models to solve complex business problems. The core focus was on the practical application of statistical methods, moving beyond theory to demonstrate how these techniques can lead to better decisions and insights. We covered a wide array of topics, starting with data visualization and distribution fitting, and progressing to predictive modeling with simple, multiple, and logistic regression. The course also introduced advanced analytical frameworks like optimization for profit maximization, Data Envelopment Analysis for efficiency measurement, and Conjoint Analysis for understanding customer preferences.

Instructor

Prof. Rahul R Marathe, Department of Management Studies, IIT Madras

Course Schedule & Topics

The course is structured to cover a range of statistical and analytical techniques with a strong focus on business applications.

Week(s) Primary Focus Key Topics Covered
1 Basic Data Visualization Fundamental techniques for visualizing and exploring business data.
2 Distribution Fitting & Goodness-of-Fit Guessing distributions, Chi-Squared Goodness of Fit (GoF) tests, P-P plots, and Q-Q plots.
3 Chi-Squared Tests & Naive Bayes Chi-Squared Test of Independence (TOI) and the Naive Bayes classification algorithm.
4 Demand-Response Curve & Simple Linear Regression Understanding the Demand-Response curve and building Simple Linear Regression (SLR) models.
5 Optimization for Business Profit and revenue maximization problems, and Primal-Dual conversion in optimization.
6 Multiple Linear Regression (MLR) Building MLR models, interpreting Adjusted $R^2$, interpreting ANOVA and checking for multicollinearity with Correlation & VIF.
7 Logistic Regression & Classification Logistic Regression for classification, and evaluating models using Accuracy, Recall, Precision, & Confusion Matrix.
8-10 Data Envelopment Analysis (DEA) A non-parametric method for measuring the productive efficiency of decision-making units.
11-12 Conjoint Analysis A survey-based statistical technique used to determine how people value different attributes of a product or service.

Material used