Prediction of Student Dropout in E-Learning Program Through the Use of Machine Learning Method

The dropout high rate is a serious problem in E-learning programs. Thus, it is a concern of education administrators and researchers. Predicting the dropout potential of students is a workable solution for preventing dropouts. Based on the analysis of related literature, this study selected students' personal characteristics and academic performance as input attributions. Prediction models were developed using Artificial Neural Network (ANN), Decision Tree (DT) and Bayesian Networks (BNs). A large sample of 62,375 students was utilized in the procedures of model training and testing. The results of each model were presented in a confusion matrix and were analyzed by calculating the rates of accuracy, precision, recall, and F-measure. The results suggested all of the three machine learning methods were effective for student dropout prediction, but DT presented a better performance. Finally, some suggestions were made for future research.