This study shows that EDM has been employed for various purposes in the education sector but the most significant application of EDM is related to the student performance prediction. Performance of the students in the education sector is not only related to their academic performances but also very much associated with their behavior and motivations. The focus of this study is to review a comprehensive amount of available literature which is related to the performance prediction of students. While conducting this study, it was noted that most of the studies are very much focused towards attain high accuracies by employing various modeling techniques while overlooking a critical problem of class imbalance. Therefore, the issue of class imbalance has also been reviewed in this study. Moreover, it was concluded that majority of these researches have catered this issue in binary-class classification problems and there is a significant space to handle the class imbalance issue for multi-class classification problems while predicting the students` performance.
Volume 12 | 07-Special Issue
Pages: 1612-1621
DOI: 10.5373/JARDCS/V12SP7/20202265