From Data to Decisions: Predictive Machine Learning Models for Customer Retention in Banking

Abstract

This research proposes machine learning—a subfield of artificial intelligence-based approaches to forecast bank client attrition. The study promotes investigating the possibility of churn through the examination of consumer behaviour. In this paper, a method to predict customer churn in a bank using machine learning techniques, which is a branch of artificial intelligence, is proposed. The research encourages the exploration of the likelihood of churn by analysing customer behaviour. The number of service providers is increasing very rapidly in every business these days. As a result, customer churn and engagement have become one of the top issues for most of the banks. In this work, the Random Forest, SVC, XGB, LGBM, and Logistic Regression classifiers are employed. A few feature selection strategies have also been applied to validate system performance and determine which features are more relevant. The test was conducted using the churn modelling dataset from IEEE dataset. The results are compared to find an appropriate model with increased precision and predictability. As a result, when utilised after oversampling, the Random Forest model outperforms other models in terms of accuracy. The Random Forest model showed a significant improvement in accuracy compared to the other classifiers, making it the preferred choice for predicting customer churn in the banking sector. The feature selection strategies helped identify the most relevant features that contribute to accurate predictions. Overall, this study highlights the importance of implementing advanced machine learning techniques and feature selection methods to enhance customer retention strategies in the banking industry.