Predictive Modeling for Student Performance Analytics Through Data Mining Techniques

Abstract

The paper attempts to predict the performance of the students, through experimental analysis and predictive modeling. There is a lack of predictive studies and models used in the Ethiopian context to accurately determine the influencing factors of the students’ performance in academics by categorizing student status into dropout/fail, poor, good, excellent or average performer. Many educational institutions have still no strategic plan to predict or determine the student’s performance in order to improve it, reduce dropouts and help to implement the curriculum/academic policies based on student’s performance and status. The study aims at conducting a comparative analysis and predictive modeling for knowing the student’s performance status through data mining techniques. The study uses the KDD process model and interpret patterns in repositories. Decision tree (J48 and Random Forest), Bayes (NaiveBayes and BayesNet) and Rule-Based (JRip and PART) algorithms are used for classification. The results reveal that the overall accuracy of the tested classifiers is above 80%. In addition, classification accuracy for the different classes reveals that the predictions are worst for fail class and fairly good for the average class. The J48 and JRip classifiers relatively produce the highest classification accuracy for the average performer/ status. Finally, the study suggests that data mining can be used as a significant technique to figure out student’s performance based on salient factors affecting.