Student Performance Prediction

This project analyzes the Student Performance dataset (UCI Repository) to understand factors affecting academic success. We applied data cleaning, EDA, clustering (K-Means), and supervised learning (Logistic Regression, Decision Tree, Random Forest) to predict whether a student will pass or fail.

The goal is to:

Identify key risk factors for student failure.

Compare model performance with and without data leakage (G1 & G2).

Provide actionable recommendations for early interventions.