Multiple Regression & Diagnostics: NSSO & IPL Analysis

Application of multiple linear regression and diagnostic techniques to analyze consumption patterns and the relationship between performance and salary using real-world datasets.

---

Overview

This project applies regression modeling and diagnostic testing on two datasets to evaluate model performance, validate statistical assumptions, and derive meaningful insights.

It combines econometric analysis with real-world applications in consumption behavior and sports analytics.

---

Part 1: NSSO Consumption Analysis

Techniques Used

• Multiple Linear Regression
• Regression Diagnostics
• Ridge Regression

Objective

To model total consumption using various food consumption variables and evaluate the validity of the regression model.

Key Analysis

• Built regression model using multiple predictors

• Evaluated assumptions:

  • Normality (Shapiro-Wilk test)
  • Homoscedasticity (Breusch-Pagan test)
  • Linearity & independence (residual analysis)

• Applied GVLMA for overall validation

Key Insights

• Extremely high R² (~1.0) indicates possible overfitting or multicollinearity
• Residuals do not follow normal distribution
• Presence of multicollinearity among predictors
• Ridge regression used to improve model stability

---

Part 2: IPL Performance vs Salary Analysis

Techniques Used

• Multiple Linear Regression
• Residual Diagnostics

Objective

To analyze the relationship between player performance and salary in the IPL.

Key Analysis

• Modeled salary (Final Price) using:

  • Runs
  • Wickets

• Evaluated model using residual plots and diagnostics

Key Insights

• Player performance significantly impacts salary
• Both runs and wickets contribute positively to salary prediction
• Model assumptions require refinement for better reliability

---

Dataset

This project uses two datasets:

• NSSO Consumption Dataset Contains detailed consumption data across regions and commodities in India Used for regression modeling and diagnostic analysis

• IPL Dataset Includes player performance metrics and salary information Used to analyze the relationship between performance and compensation

---

Applications

• Consumption pattern analysis
• Economic modeling
• Sports analytics and salary prediction
• Model validation and diagnostics

---

Tech Stack

• Python (pandas, statsmodels, matplotlib)
• R (lm, glmnet, gvlma)
• Statistical modeling and econometrics

---

Project Structure

• analysis.ipynb – Python implementation
• analysis.R – R implementation
• README.md – project overview
• project_report.pdf – detailed analysis
• data – datasets