This project focuses on predicting customer churn in a retail bank using machine learning techniques. Churn — when customers leave the bank — directly impacts profitability. By analyzing customer demographics, account information, and behavioral patterns, this project aims to identify the most influential features contributing to churn and build predictive models to assist in retention strategies.
Age
Gender
Balance
Credit Score
Tenure
Number of Products
Geography
Estimated Salary
Churn Label (Exited/Stayed)
Python (Jupyter Notebook)
Pandas, NumPy – Data manipulation & preprocessing
Matplotlib, Seaborn – Visualization
Scikit-learn – Machine learning models & evaluation
Imbalanced-learn – Handling class imbalance
Data Exploration & Cleaning
Checked missing values, duplicates, and outliers
Handled categorical encoding and scaling
Exploratory Data Analysis (EDA)
Visualized churn distribution and key customer features
Identified relationships between churn and variables (e.g., age, balance, geography)
Feature Engineering
Encoded categorical features
Normalized/standardized numerical features
Trained multiple ML models (Logistic Regression, Random Forest, XGBoost, etc.)
Evaluated with accuracy, precision, recall, F1-score, and ROC-AUC
Insights & Recommendations
Highlighted most important features driving churn
Suggested strategies for customer retention