The Development of Student Academic Performance Prediction System for UPTM using RepTree Algorithm

Abstract

Accurate prediction of students' performance is critical in this digital age for educational institutions to identify at-risk pupils and provide timely interventions. Numerous models have been proposed under different educational contexts to address it, but there is a lack of sophisticated models causing difficulty for the user in giving guidance to the stakeholders to take appropriate measures to counter student’s problems. The RepTree algorithm is a decision tree-based machine learning approach to modelling and forecasting student academic achievement. This study adopted it to develop a Student Academic Performance Prediction System (SAPPS) for the University Poly-Tech Malaysia (UPTM). This study is using a mixed-method research design, based on historical student data to train the predictive model, including demographics, prior academic performance, and sponsorship attributes. The evaluation result shows how well the established system predicts students' academic success and how reliable it is. The system's predictive capabilities give educational institutions the ability to spot students who are likely to have academic difficulties and take proactive measures to improve their results. The use of such a predictive system could improve the learning environment and boost students' achievement at UPTM. By having this information, the target users like administrators, lecturers, and academic advisers can access and decipher the predictions produced by the RepTree algorithm using the suggested system's user-friendly interface.