Student grouping by neural network based on affective factors in learning English

Differences between teaching perspectives and students characteristics may impact negatively on students' learning effectiveness. A new approach to bridge such a gap needs establishing. The capabilities of artificial neural networks to approximate extremely complex problems encourage us to develop a grouping model of students' English ability. The model was trained using back propagation algorithm and tested using 154 samples from college students. The model grouping rate on students' English abilities demonstrated fairly low errors for both general grouping and each ability grouping for Listening, Reading, Speaking, and Reading, respectively.