After outputting the results of the model, the infuence of the input features on the output result should be analyzedreasonably. SHAP, as an explanation method of AI models, has beenapplied in many felds with good results, indicating machine learningprediction (e.g., XGBoost, LightGBM, etc.) is frst taken as an example