Research on Prediction Bank Credit Card User Churn Based on XGBoost

To improve the service experience of bank credit card customers more accurately and reduce operating costs, this paper predicts the churn risk of bank credit card users. Based on the analysis of bank credit card customer consumption data released by the Kaggle platform, this paper uses SMOTE data enhancement technology to construct a bank churn customer prediction model based on XGBoost and compared the effectiveness of logistic regression, decision tree, and random forest machine learning algorithms. The results show that the XGBoost algorithm shows the best prediction accuracy, reaching 95%, which provides strong data support for banks in retaining credit card customers.