Customer Churn Prediction Using Synthetic Minority Oversampling Technique

Abstract

Customer churn prediction is the process of using data and analytics to forecast that customers are likely to stop using a product or service. It helps businesses proactively identify and retain at-risk customers by implementing targeted retention strategies. Business analysts and CRM (Customer Relationship Management) analysts should comprehend the effects behind customer churn and analyze the behaviour trends of existing churn customers from the data. From telecom sector, gaining new consumers is not considered as a smart move because it costs far less to keep existing ones. This approach combines customer segmentation and churn prediction to give telecom operators a comprehensive customer churn analysis. It helps in businesses reduce costs, retain customers through personalized interventions, and gain a competitive edge by leveraging data-driven insights to improve overall profitability and customer satisfaction. Four machine learning (ML) classifiers are employed in the experiments. First, these ML classifiers are used to predict the customer's churn state. To focus on the imbalanced datasets issues, we apply the proposed Synthetic Minority Oversampling Technique (SMOTE). The best method is Gradient Boosting Classifier, which has achieved an accuracy of 95.13 percent, based on the experimental investigation.