<![CDATA[The rapid growth of e-learning platforms has resulted in an enormous amount of student interaction data, creating opportunities to anticipate learning outcomes and implement timely interventions. In this research, a Stacked Classifier Model (SCM) is introduced to predict student performance using e-learning reaction data obtained from a Kaggle repository. The SCM employs a hierarchical ensemble approach by combining several base classifiers—K-Nearest Neighbors (KNN), Decision Tree (DT), Multi-Layer Perceptron (MLP), Random Forest (RF), Support Vector Machine (SVM), and Radial Basis Function (RBF) networks—to capitalize on their respective strengths while compensating for individual limitations. The dataset underwent careful preprocessing, including imputation, encoding, feature normalization, and temporal aggregation, to ensure the classifiers received high-quality input. Evaluation results indicate that the SCM outperforms each base model individually, demonstrating its capability to capture complex behavioral patterns in e-learning contexts. Overall, this study highlights the effectiveness of ensemble learning techniques in educational data mining, offering a solid foundation for adaptive learning, personalized interventions, and enhanced academic performance.]]>
Copyrights © 2026
