PREDICTIVE ANALYTICS FOR CUSTOMER CHURN IN BANKING: A MACHINE LEARNING APPROACH TO RETENTION

Abstract

Using cutting-edge AI methods focused on machine learning models like Random Forest, XGBoost, and Logistic Regression, this study investigates the prediction of customer attrition in the banking industry. The highest accuracy of 81.07% is achieved from Logistic Regression. XGBoost achieves 80.31% with similar results. The result of feature importances analysis shows that TotalCharges, MonthlyCharges and Contract have the highest influence on churn. Further clustering breaks customers down into an actionable group for targeted retention strategy. The results indicate that predictive analytics can be used to reduce churn, boost customer satisfaction, and improve the performance of the business, thus requiring data-based customer retention strategies in banking.