Predicting Banking Customer Churn Using Machine Learning

Customer churn is a major challenge in the banking industry, as customers may stop using a bank’s services and move to other banks for various reasons. Identifying customers who are likely to leave is important for banks to improve customer retention and maintain long-term customer relationships. The project titled Banking Customer Churn Prediction Using Machine Learning aims to analyse customer data and predict whether a customer is likely to stay with the bank or leave. The system begins by preprocessing the dataset to prepare the data for analysis, followed by Exploratory Data Analysis (EDA) to understand customer behaviour, identify patterns, and examine the factors associated with customer churn. Customer information, such as age, credit score, account balance, tenure, number of products, geographical location, and account activity, is analysed based on the available dataset. Machine learning classification algorithms are trained and evaluated to identify a suitable model for predicting customer churn. The selected XGBoost model is used to estimate the probability of a customer leaving the bank. A user-friendly dashboard is developed using Streamlit to present the analysis and prediction results through graphs, a correlation heatmap, customer risk distributions, geographical comparisons, and individual customer predictions. The dashboard also allows users to enter customer details and view the predicted churn probability, helping them understand whether a customer may be at risk of leaving. The proposed system helps identify customers who may require additional attention and provides useful insights into customer behaviour. These insights can support banks in developing appropriate customer retention strategies and making informed business decisions. Thus, the project demonstrates the practical application of Machine Learning and Data Analytics in predicting banking customer churn and supporting effective customer relationship management.