Customer Churn Prediction using Confident Learning and XGBoost

Customer churn prediction is a critical task for businesses to reduce customer attrition and improve customer retention strategies. In this paper, we present a method to enhance the accuracy and effectiveness of churn prediction in the banking sector by utilizing confident learning techniques alongside the XGBoost algorithm. Confident learning is employed to address the challenges of uncertain or mislabeled training data, while XGBoost is utilized as a powerful classification algorithm for building an efficient prediction model. The experimental results showed a significantly improved accuracy from 76.8% to 81.2%.