Single Classifiers and Ensemble Approach for Predicting Student’s Academic Performance

International Journal of Research and Scientific Innovation (IJRSI) | Volume VII, Issue VI, June 2020 | ISSN 2321–2705

Single Classifiers and Ensemble Approach for Predicting Student’s Academic Performance

OLUKOYA, Bamidele Musiliu
M.Sc Student, University of Ilorin, Nigeria

Abstract: In recent time, educational data mining (EDM) has received substantial considerations. Many techniques of data mining have been proposed to dig out out-of-sight knowledge in educational data. The Knowledge obtained assists the academic institutions to further enhance their process of learning and methods of passing knowledge to students. Consequently, the performance of students soar and the educational products are by no doubt enhanced. In this study, a novel student’s performance prediction model premised on techniques of data mining with Students’ Essential Features (SEF). Students’ Essential Features (SEF) are linked to the learner’s interactivity with the e-learning management system. The performance of student’s predictive model is assessed by set of classifiers, viz. Bayes Network, Logistic Regression and REP Tree. Consequently, ensemble methods of Bagging Boosting and Random Forest are applied to improve the performance of these single classifiers. The results obtained reveal that there is a robust affinity between learner’s behaviors and their academic attainment. Results from the study shows that REP Tree and its ensemble record the highest accuracy of 83.33% using SEF. Hence, in terms of Receiver Operating Curve (ROC), boosting method of REP Tree records 0.903, which is the best. This result further demonstrates the dependability of the proposed model.

Keywords: EDM, Ensemble, Bagging, Boosting, Random Forest, Data mining, Classifiers, machine learning.

I. INTRODUCTION

There have been tremendous changes in educational setting across the globe in its functioning (Mishra, 2014). The increase of e-learning resources, instrumental educational software, the use of the Internet in education, and the establishment of state databases of student information has created large repositories of data (Koedinger, 2008). All this information provides a goldmine of educational data that can be explored and exploited to understand how students learn (Mostow, 2006). In fact, today, one of the biggest challenges that educational institutions face is the exponential growth of educational data and the use of this data to improve the quality of managerial decisions (Bala, 2012).

Educational data mining (EDM) is concerned with developing, researching, and applying computerized methods to detect patterns in large collections of educational data that would otherwise be hard or impossible to analyze due to the enormous volume of data within which they exist (Romero, 2010). EDM has emerged as a research area in recent years aimed at analyzing the unique kinds of data that arise in educational settings to resolve educational research issues (Baker and Yacef, 2009).

Read More