PREDICTING STUDENTS’ PERFORMANCE USING AN ENHANCED AGGREGATION STRATEGY FOR SUPERVISED MULTICLASS CLASSIFICATION | ||
| International Journal of Intelligent Computing and Information Sciences | ||
| Article 22, Volume 22, Issue 3, August 2022, Pages 124-137 PDF (886.07 K) | ||
| Document Type: Original Article | ||
| DOI: 10.21608/ijicis.2022.146420.1195 | ||
| Authors | ||
| Mohamed Farouk Yacoub* 1; Huda Amin2; Nivin Atef3; Sebastián Ventura Soto4; Tarek Gharib5 | ||
| 1Information Systems Department, Faculty of Computer and Information Sciences, Ain Shams University | ||
| 2Faculty of Computer and Information Sciences,Ain shams University | ||
| 3Faculty of computer and information sciences- Ain Shams University | ||
| 4Department of Computer Sciences and Numerical Analysis, University of Cordoba, Spain. | ||
| 5Information Systems Department, Faculty of Computer and Information Sciences, Ain Shams University, Cairo, Egypt | ||
| Abstract | ||
| Predicting students performance efficiently became one of the most interesting research topics. Efficiently mining the educational data is the cornerstone and the first step to make the appropriate intervention to help at-risk students achieve better performance and enhance the educational outcomes. The objective of this paper is to efficiently predict students’ performance by predicting their academic performance level. This is achieved by proposing an enhanced aggregation strategy on a supervised multiclass classification problem to improve the prediction accuracy of students’ performance. Two binary classification techniques: Support Vector Machine (SVM) and Perceptron algorithms, have been experimented to use their output as an input to the proposed aggregation strategy to be compared with a previously used aggregation strategy. The proposed strategy improved the prediction performance and achieved an accuracy, recall, and precision of 75.0%, 76.0%, and 75.48% using Perceptron, respectively. Moreover, The proposed strategy outperformed and achieved an accuracy, recall, and precision of 73.96%, 73.93%, and 75.33% using SVM, respectively. | ||
| Keywords | ||
| Keywords: Machine Learning; Students’ performance prediction; Educational data mining; Multiclass Classification; Supervised Learning | ||
|
Statistics Article View: 204 PDF Download: 278 |
||
