Predicting Student Dropout Rates in Online Education Platforms utilizing Naive Bayes Algorithm

High student dropout rates pose a significant challenge to the effectiveness of online education platforms. This study utilizes the Naive Bayes algorithm to predict student dropouts by analyzing data on academic performance, engagement, demographics, and behavior. Key techniques such as SMOTE and Recursive Feature Elimination (RFE) are employed to address class imbalance and enhance model accuracy. The Naive Bayes classifier is optimized using scalable machine learning frameworks, enabling real-time predictions for timely interventions. Evaluation metrics like accuracy, precision, recall, F1-score, and ROC-AUC demonstrate the model's superior performance compared to traditional methods. Results indicate that Naive Bayes effectively identifies at-risk students, providing valuable insights for targeted support strategies. This research advances educational data mining by demonstrating the practical application of Naive Bayes in reducing dropout rates and fostering a more supportive online learning environment.