Machine Learning Based Evaluation and Prediction of Student Performance in Virtual Learning Environment
In recent years, the rapid expansion of Virtual Learning Environments (VLEs)and online education platforms has significantly transformed higher education.These education setups have introduced new challenges for example increased student attrition and disengagement. Predicting student performance within these digital frameworks is essential to enabling timely interventions and improving academic outcomes. This research addresses the prediction of high risk students by leveraging the Open University Learning Analytics Dataset (OULAD) that comprises demographic, assessment, and detailed engagement data for over 32,000 students. Recognizing gaps in the literature, this study develops a robust ensemble based machine learning pipeline with Particle Swarm Optimization(PSO) as a feature selection strategy. The results demonstrate that PSO reduced the feature space by approximately 40% while maintaining high predictive performance (F1 ≈ 0.937, AUC ≈ 0.980) in comparison to Gini importance and Factor Analysis of Mixed Data(FAMD). Overall 94% accuracy and AUC scores near 0.981 is achieved that is comparable to advanced deep learning benchmarks. The inclusion of SHAP analysis improves the interpretation of the proposed scheme. This work contributes a balanced, explainable framework for educational early warning systems thereby, bridging methodological gaps in comparative feature selection techniques and offering practical insights for scalable deployment in online learning contexts.