Blended Learning Behaviors Analysis Based on BP-Bagging Classification Model

To address the challenges of monitoring and analyzing blended learning behaviors, a BP-Bagging ensemble classification model to discriminate students’ blended learning behaviors is proposed. Considering the multiple features and complex interrelationships of blended learning behaviors, the PI algorithm is employed to extract learning behavior features, and the BP neural network is combined with the Bagging algorithm to construct an ensemble model for classifying and discriminating students’ learning behaviors. Experimental results demonstrate that the BP-Bagging ensemble model achieves an accuracy rate of 93.6% in discriminating blended learning behaviors. Compared with other classification models, this ensemble model can improve the generalization performance of individual BP neural network models and has significant practical value.