Ensemble of Machine Learning Algorithm for Customer Churn in Telecom Industry

Abstract

Churn occurs when customers switch providers due to dissatisfaction or competitor offers, causing major losses since retention is cheaper than acquisition. With rising global competition, customer retention is vital for sustainability. Telecom companies now analyze massive CDR data to detect churn patterns and predict likely churners early. This enables effective retention strategies, service improvements, and targeted marketing. The proposed model preprocesses CDR data, rebalances it, applies machine learning for classification, and uses ensemble learning to enhance accuracy and generate reliable churn insights.