A Boosting Machine Learning Model for Predicting Student Performance in E-Learning

In the wake of the COVID-19 pandemic, e-learning systems have gained renewed importance. Researchers have focused on improving these platforms by addressing key challenges in MOOCs, with student performance being one of the most critical areas. The widespread use of e-learning platforms has generated vast amounts of valuable data, providing rich opportunities for data mining research. In this study, we aim to predict student performance and dropout using data from the OULAD dataset. The Boosting method was applied, utilizing machine learning models such as Linear Discriminant Analysis (LDA), Decision Trees (DT), Random Forest (RF), Support Vector Machine (SVM), Naive Bayes (NB), and Logistic Regression (LR). After experimenting with various approaches, Random Forest achieved the best performance with F1-scores of 86%, 81%, and 93% for the Dropout, Fail, and Pass classes, respectively.