A FUZZY LOGIC APPROACH TO PREDICTING STUDENTS’ MATHEMATICS PERFORMANCE

Abstract

Mathematics performance plays a vital role in academic success; however, recent trends indicate a decline, especially among science students. Three factors that may contribute to the students’ performance in Calculus I are attendance, test scores, and study hours. This study proposes a predictive model based on Fuzzy Logic to assess the mathematics performance of Universiti Teknologi MARA (UiTM) students in Calculus I. Developed using MATLAB, the model integrates three primary input variables; attendance, study hours, and test scores are selected for their significant impact on academic achievement as established in prior studies. Data was collected from students in the Faculty of Applied Sciences (FSG) and the College of Computing, Informatics, and Mathematics (KPPIM) at the UiTM Perlis Branch. The model was validated and found to reliably predict student performance. Among the variables, test scores emerged as the most significant predictor, highlighting their potential as an early indicator for identifying students at risk of underperforming. These insights can assist lecturers in implementing timely, targeted interventions to support academic improvement. In conclusion, the Fuzzy Logic model provides an effective, data-driven approach for predicting mathematics performance. It offers valuable support for educators seeking to enhance student outcomes in mathematics-related courses.