Achieved 87% R² score using CatBoost, enabling precise predictions of student math performance.
Revealed socioeconomic factors (lunch type) and test preparation as the strongest performance predictors.
Successfully implemented end-to-end CI/CD pipeline for Azure Web App deployment with Docker containerization.
Early Intervention Programs
Resource Allocation Optimization
Personalized Learning Paths
Data-Driven Policy Making
Predictive analytics in education can help close achievement gaps by identifying at-risk students early and allocating resources more effectively, potentially improving graduation rates by 15-20% when combined with targeted intervention programs.