A study of multi-factors affecting college students'learning performance——Based on interpretable machine learning model
This paper added SHAP interpretation method on the basis of machine learning model to conduct an in-depth study on the potential factors,which include the requirements of prediction accuracy and inter-pretation in the actual situation.The results show that personal factors and school education factors of col-lege students have great influence,while family factors have relatively little influence.It can be seen that gender,age,extra work,partner,and lectures have a significant impact from the results of SHAP based on GBDT model.It is necessary for the university to intervene in advance from four dimensions:guiding the failure anxiety of the university students,guiding the university students to establish the correct view of love,helping the university students to establish the concept of lifelong learning and paying attention to the practical and innovative ability of the university students so as to improve the academic performance of the university students.
万方数据
