Using Decision Tree and Artificial Neural Network to Predict Students Academic Performance

Abstract

Student Academic Performance is a great concern for academic institutions in all levels of academic years. Techniques like classification, clustering and association are provided by Data Mining. In this paper, two classification techniques, Decision Tree (J48) and Artificial Neural Network (ANN), are used to build a classification model, that can predict the academic performance of university students in Jordan, expected GPA in precise. A dataset has been gathered using online questionnaire, and certain attributes were selected to test their relevance to the academic performance of a Jordanian university students. The paper describes the methodology conducted to apply the J48 and ANN, using a special tool (WEKA), and the results are discussed in details, showing a better performance for ANN in some cases, and a better performance for Decision Tree in others.