Predicting students academic performance using artificial neural network

Description

Abstract University students’ retention and performance in higher education are important issues for educational institutions, educators, and students. The observed poor quality of graduates of some Nigerian Universities in recent times is traceable to non-availability of mechanism that would enable the University administrators to project into the future performance of the concerned students. This will guarantee the provision of better educational services as well as customize assistance according to students’ predicted level of performance. Educational data mining is focused on developing models and methods for exploring data collected from educational environments for better understanding and improvement of the educational process. Analyzing and determining patterns among indicators of academic success (study grade point average) and their correlation to students’ personal, high school, admission data can be a good foundation in the process to adapt and improve the curriculum of higher education institutions, according to the students’ characteristics. Additionally, we identified which factors had a crucial influence on overall students’ success. In this research, Artificial Neural Networks (ANNs) were used to develop a model for predicting the performance of students before graduation. The data used in this study consists of eighty (80) randomly selected students in four different schools that are present in 200, 300, and 400 levels in Emmanuel Alayande College of Education Oyo (affiliation with Ekiti State University Ado-Ekiti). Test data evaluation showed that the ANN model is able to correctly predict students’ performance at 92.7% accuracy.