Analysis of Students’ Psychological Behavioral Changes in Intelligent Data Processing Environment in Mental Health Education

Mental health education is an effective means to understand students’ psychological behavior, and there are problems such as inaccurate data characterization and prediction of behavioral changes in the intelligent data processing environment, which is not conducive to the development of students’ psychological health. In order to solve the above problems, this paper proposes a prediction model of students’ psychological behavior change based on XGBoost algorithm, which uses Pearson correlation coefficient and regression analysis to screen the characteristics of students’ psychological behavior change, then establishes a prediction model based on XGBoost, and introduces the SHAP model to analyze the predictive results for interpretability. The results show that the model’s RMSE mean value is only 2.454 when predicting students’ psychological behavior changes, and the model’s fitting effect on the dataset reaches 0.8924. Relying on the results obtained by the model, the characteristics of students’ psychological behavior changes can be clarified, which can provide guiding references for the implementation of mental health education.