Customer Churn Prediction in the Banking Sector Using Machine Learning-Based Classification Models

Aim/Purpose: Previous research has generally concentrated on identifying the variables that most significantly influence customer churn or used segmentation to identify a subset of potential consumers, excluding its effects forecast accuracy. Consequently, there are two primary goals in this work. The initial goal was examine impact accuracy prediction banking sector using machine learning models. second objective is experiment, contrast, and assess which approaches effective predicting churn. Background: This paper reviews theoretical basis churn, segmentation, suggests supervised machine-learning techniques for attrition prediction. Methodology: In study, we use different models such as k-means clustering segment customers, k-nearest neighbors, logistic regression, decision tree, random forest, support vector apply dataset predict Contribution: results demonstrate performs well with forest model, an about 97%, that, following mean each model performed well, regression having lowest (87.27%) best (97.25%). Findings: Customer does not have much precision predictions. It dependent choose. Recommendations Practitioners: practitioners can proposed solutions build predictive system them other fields education, tourism, marketing, human resources. Recommendation Researchers: paradigm also applicable areas artificial intelligence, learning, Impact Society: will cause value flowing from customers enterprises decrease. If continues occur, enterprise gradually lose competitive advantage. Future Research: Build real-time near application provide close information make good decisions. Furthermore, handle imbalanced data new techniques.