Applied machine learning
Student Grade Prediction System
Built an early-risk screening workflow from 30 start-of-term features, deliberately excluding prior grades. The app compares six classifiers, tunes Random Forest and XGBoost with GridSearchCV, and exposes prediction, ROC, model comparison, and feature-insight views.