<![CDATA[Bullying behavior is a serious problem in the educational environment that has an impact on students' psychological health and social development. Psychosocial factors such as empathy, emotion regulation, peer pressure, and school climate are known to contribute to the emergence of these behaviors. This study aims to develop a model of predicting bullying behavior using a Machine Learning approach based on psychosocial factors and evaluate the accuracy level of the resulting model. This study used a predictive quantitative design with a cross-sectional approach on 412 junior high and high school students. Data was collected through a standardized questionnaire and analyzed using four classification algorithms, namely Logistic Regression, Support Vector Machine (SVM), Random Forest, and XGBoost. Model validation was carried out using a 5-fold cross validation technique with evaluation parameters of Accuracy, Precision, Recall, F1-score, and ROC-AUC. The results showed that the XGBoost algorithm had the best performance with an Accuracy value of 90%, F1-score of 0.85, and a ROC-AUC of 0.93. The variables of peer pressure, empathy, and emotion regulation were the most influential predictors in the model. These findings suggest that the integration of psychosocial factors and Machine Learning techniques is effective in building accurate models of predictive bullying behavior. This model has the potential to be used as a basis for the development of early detection systems and preventive interventions in the school environment]]>
Copyrights © 2025
