Modeling Dropout Behavior Patterns Using Bayesian Networks in Small-Scale Private University

Abstract

In this study, we combine the whole student model and the individual student model to express three kinds of dropping behavior, heterogeneity of the dropping behavior and temporal heterogeneity. The model constructed in this research is still a skeleton model, In the future, quantitative survey, statistical learning should be conducted, learning and expression of graph structure and quantitative relation are required.