KMAN Publication Inc. (KMANPUB) Journal of Assessment and Research in Applied Counseling (JARAC) 8 Serial Number 30 2026 07 01 Support Vector Machine Classification of High-Risk University Students Using Academic Stress, Sleep Quality, Self-Efficacy, and Depression Support Vector Machine Classification of High-Risk University Students Using Academic Stress, Sleep Quality, Self-Efficacy, and Depression 1 14 EN Journal Article 2026 03 19 <table> <tbody> <tr> <td> <p><strong>Objective:</strong> This study aimed to develop and evaluate a Support Vector Machine classification model for identifying high-risk university students in Colombia using academic stress, sleep quality, self-efficacy, and depression as predictive indicators.</p> <p><strong>Methods and Materials:</strong> This quantitative cross-sectional predictive study was conducted among 1,248 undergraduate students from five public and private universities in Colombia. Data were collected using standardized self-report instruments measuring academic stress, sleep quality, general self-efficacy, and depressive symptoms. Students were classified into high-risk and lower-risk groups according to predefined criteria combining depressive symptoms, severe academic stress, and poor sleep quality. Data were analyzed using descriptive statistics, correlation analysis, and a Support Vector Machine algorithm with a radial basis function kernel. The dataset was divided into training and testing subsets using stratified sampling, and hyperparameter optimization was performed through grid search and ten-fold cross-validation. Model performance was evaluated using accuracy, precision, recall, specificity, F1-score, balanced accuracy, Matthews correlation coefficient, ROC-AUC, and Brier score. SHAP analysis was applied to interpret the relative contribution of each predictor.</p> <p><strong>Findings:</strong> Correlation analysis showed that academic stress was positively associated with depression and poor sleep quality, while self-efficacy was negatively associated with academic stress, poor sleep quality, and depression. The optimized Support Vector Machine model demonstrated strong classification performance, achieving 92.4% testing accuracy, 90.6% precision, 89.1% recall, 93.7% specificity, an F1-score of 89.8%, balanced accuracy of 91.4%, Matthews correlation coefficient of 0.81, ROC-AUC of 0.951, and Brier score of 0.074. SHAP analysis identified depression as the strongest predictor of high-risk classification, followed by academic stress, sleep quality, and self-efficacy.</p> <p><strong>Conclusion:</strong> The findings indicate that Support Vector Machine classification can accurately and interpretably identify high-risk university students based on psychological and behavioral indicators. Integrating depression, academic stress, sleep quality, and self-efficacy into predictive screening may support early detection and targeted student mental health interventions.</p> </td> </tr> </tbody> </table> https://journals.kmanpub.com/index.php/jarac/article/download/5482/10962