Ensemble-Based Machine Learning Strategies for Predicting Student Dropout and Academic Outcomes.

This paper presents an ensemble-based machine learning framework for predicting student dropout and academic success in higher education. Using real-world institutional data comprising demographic, socio-economic, and academic performance variables, multiple supervised learning models are evaluated, including Logistic Regression, Decision Tree, SVM, KNN, Random Forest, and a Stacking Classifier. Experimental results demonstrate that the ensemble approach achieves superior predictive performance, highlighting the effectiveness of data-driven decision support systems for improving student retention and academic outcomes.