This project evaluates and compares GLM and LASSO logistic regression models for predicting depression risk in ~27,900 students. The study emphasizes interpretability, cross-validated performance, and feature-level analysis to identify robust psychological and lifestyle predictors