Predictive modeling of student success using machine learning

This paper presents a comprehensive study of predictive models for student success, leveraging machine learning techniques to analyze diverse factors affecting academic performance in higher education. Using a dataset of students from various institutions in Kosovo, encompassing demographic, socioeconomic, and academic attributes, the research employs and evaluates several machine learning algorithms to predict the dropout rate. To address the issue of unbalanced data, SMOTE and ADASYN techniques were implemented, enhancing the predictive accuracy. The study identifies significant predictors of student success, and the results indicate that machine learning can accurately predict student outcomes, providing educational institutions with actionable insights to mitigate dropout rates and enhance student support.