Engineering Students’ Performance Prediction on Board Examination Using Classification Algorithms

Board examinations are critical assessments that determine the academic and professional readiness of engineering students. Accurately predicting board exam outcomes can support timely interventions, helping institutions and educators enhance student preparedness. This study developed a predictive model using machine learning classification algorithms, specifically logistic regression, decision trees, random forest, and Naïve Bayes, to forecast the board examination performance of engineering students based on academic and preparatory indicators such as general weighted average, pre-board scores, and review center participation. Among the models tested, logistic regression achieved the highest accuracy (66.7%), closely followed by Naïve Bayes (66.1%). The findings emphasize the predictive value of pre-board performance and institutional review programs. This research highlights how predictive analytics can improve educational strategies and support systems, ultimately aiming to raise board exam success rates. Future research is encouraged to integrate additional variables, including psychological and behavioral factors, to further enhance model accuracy.