Machine Learning on Retail Stores for Customer Churn

The pace at which customers discontinue engaging in commercial activity with a corporation is known as customer churn, and it frequently indicates discontent or competition. Machine learning analyzes customer data to identify patterns, predict churn, and personalize retention strategies. It enables businesses to target at-risk customers and enhance engagement through tailored interventions and offers. A public dataset gathered from Kaggle is used in this research work. Exploratory data analysis is carried over to understand the patterns and trends in the churn rate. The bivariate analysis is held to find the correlation of each feature against the churn rate. With these observations, the importance of marketing strategies is exhibited. Additionally, four distinct models—Logistic Regression, Random Forest, Decision Tree, and the XGB Classifier-are used to train and assess the dataset. From the evaluation of all four models with the help of the ROC score, the XGB is found to outperform and provide better results obtaining 98% accuracy and a 1.00 ROC score.