Abstract
<jats:p>This research investigates the analysis and prediction of student behavior in virtual learning platforms using machine learning methods. In recent years, the data collected from online learning environments regarding students' learning activities has enabled the analysis of their behaviors and the development of personalized teaching strategies. The study applies machine learning models such as Decision Tree, Random Forest (RF), Support Vector Machine (SVM), and k-Nearest Neighbors (k-NN), evaluating the impact of each model on student outcomes and its predictive capabilities. The results show that the RF model provides more accurate and reliable results, while the k-NN model performs poorly with large datasets and uneven distributions. The research demonstrates the effectiveness of machine learning for analyzing student behavior and indicates that more complex models can be applied in the future.</jats:p>