Construction of Heterogeneous Dynamic Grouping Pattern Based on Neural Network

Regardless of online or offline learning, there are operational difficulties in “facing all students”, and it is very difficult to pay attention to the “individual differences” between students. As we all know, students are developing people. During the teaching process, students' mentality, knowledge, and abilities will change, which may shouldn't be taken into account by static grouping. In this study, neural network model was used to construct the mapping relationship between students' characteristics and heterogeneous grouping, and the trained model was used to predict the grouping position at the next moment. This dynamic grouping algorithm can ensure that every student is in a heterogeneous group for a long time. Finally, we propose a heterogeneous dynamic grouping teaching pattern.