International Journal of
Information and Education Technology

Abstract—The integration of machine learning techniques into the educational landscape has opened new avenues for analyzing and improving learning experiences. This research investigates the predictive capability of classification algorithms in determining how external conditions affect student preferences for online learning. By analyzing variables such as internet Disruption, device availability, and psychological stress during the COVID-19 pandemic, we developed several classification models to uncover the patterns driving these preferences. The study applied a range of supervised learning algorithms—namely, random forest, logistic regression, gradient boosting, support vector machines, k-nearest neighbors, naïve Bayes, and decision trees—to identify the most accurate predictive approach. Among the evaluated classification algorithms, K-Nearest Neighbors achieved the highest accuracy (0.798) and F1-score (0.883), with strong recall (0.979). Support Vector Machine obtained the highest recall (1.000) but had a lower ROC-AUC (0.409). Logistic Regression, Naive Bayes, Random Forest, and Gradient Boosting showed balanced performance, with F1-scores ranging from 0.844 to 0.860. Decision Tree yielded the lowest accuracy (0.712) but maintained competitive precision (0.824). Overall, K-Nearest Neighbors and Support Vector Machine demonstrated superior recall, while K-Nearest Neighbors provided the best overall classification performance. To gain deeper interpretability of feature contributions, we employed SHAP (SHapley Additive exPlanations), which highlighted stress as the most influential factor. The findings offer actionable insights into how non-academic influences shape learning modality choices, supporting data-driven strategies to adapt online education to diverse student needs during crisis conditions and beyond.

Keywords—classification algorithms, virtual education, learner decision-making, external influences, predictive modeling, Learning Management Systems (LMS) platforms, educational data analytics

[PDF]

Cite: Bagher Javadi, Suwimon Kooptiwoot, Chaisri Tharasawatpipat, Sivapan Choo-in, Pantip Kayee, and Duongdearn Suwanjinda, "Evaluating Classification Machine Learning Models for Identifying External Factors Influencing Student Choices in Virtual Learning Environments," International Journal of Information and Education Technology, vol. 16, no. 1, pp. 278-287, 2026.

Copyright © 2026 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).