CUSTOMER CHURN PREDICTION USING MACHINE LEARNING

Abstract

Customer churn, where users stop using a service, is a major challenge, especially in telecom. Predicting churnhelps businesses retain customers more cost-effectively than acquiring new ones. This study uses machine learningmodels—SVM, Random Forest, and Decision Trees—to analyze customer data. Random Forest performs best inaccuracy, precision, recall, and F1-score. Integrating these models into CRM systems helps identify at-riskcustomers early. Targeted retention strategies reduce churn and improve customer loyalty. Machine learningenhances consumer behavior prediction. The study provides key insights into churn dynamics.