| A Hybrid Metaheuristic-Optimized Ensemble Voting Framework for Enhanced Student Academic Performance Prediction | |
|
MSRDG International Journal of Computer Scientific Technology & Electronics Engineering
© 2026 by MSRDG IJCSTEE Journal Volume 2 Issue 3
Year of Publication: 2026 |
Paper Download Article ID MSRDG-IJCSTEE-V2I3P105 |
|
Abstract: Predicting student academic performance in advance enables timely intervention, personalized support, and improved retention outcomes in educational institutions. However, raw educational datasets are often high-dimensional and noisy, containing many features with limited predictive value, which can degrade the accuracy of downstream classifiers. This paper proposes a hybrid framework that combines metaheuristic-optimized feature selection with a multi-model ensemble voting network to enhance student academic performance prediction. The framework first applies a bio-inspired optimization algorithm to identify an optimal subset of academic, behavioral, and demographic features, then trains multiple deep learning and ensemble base learners in parallel on the reduced feature set, and finally fuses their predictions through a weighted voting mechanism to produce the final performance classification. [Summarize your key results here once evaluation is complete, e.g., the accuracy improvement achieved by the proposed voting framework over individual base learners.] The proposed approach is evaluated using standard classification metrics — accuracy, precision, recall, and F1-score — and is designed to generalize across institutional datasets of varying size and feature composition. |
|
| Keywords: Student Performance Prediction; Feature Selection; Metaheuristic Optimization; Ensemble Learning; Voting Classifier; Deep Learning; Educational Data Mining | |
