Student's Academic Performance Modelling and Prediction: A Fuzzy Based Approach

Abstract

In higher Institutions of learnings, importance is placed on the quality of students admitted as this has direct effect on the quality of graduates been produced by the Institutions, and thus affect the national man-power quality at large. One of the challenges facing the Universities is admitting on merits and surprisingly, the academic performance of the students admitted on merit begins to drop. Therefore, it is important to predict academic performance of the students early enough so as to help instructors take appropriate action in adjusting teaching style and improve greatly on Students' success. In this paper, a fuzzy logic model is used to model data of students and predict their academic performance. Factors like student Ordinary level (O' level) grades, motivation to study in their given course and parents' academic background were used to predict students' academic success level prior to their end of first academic session. The results when compared with the actual result for the semester examination show 75% accuracy. This early academic performance prediction serves as a guide to the instructors. A good understanding of the students help the instructors to take appropriate steps for effective teaching and learning, and thus improve students' academic performance. Hence, the percentage of students withdrawn from the University after their first academic session due to poor performance (cumulative grade point average below 1.5) is expected to be reduced