Predicting Student Performance to Boost Educational Outcomes: The Efficacy of a Random Forest Approach

Predicting student performance is crucial for educational institutions aiming to enhance student achievement and academic readiness. This research, conducted at a university, utilizes machine learning algorithms, covering data from 2019 to 2022, to forecast academic performance. The study evaluates multiple regression models, with Random Forest Regression standing out with an R-squared score of 0.961. It emphasizes the importance of feature selection, identifying total units and subject grades as key predictors. These insights are essential for refining educational strategies and enabling targeted interventions and resource allocation to improve learning experiences and academic success directly. The findings provide a framework for applying educational analytics in other institutions. Future research will explore a broader range of characteristics, temporal analysis, and hybrid models, aligning with the needs of contemporary educational systems for data-driven, personalized learning.