Customer Churn Prediction in Telecommunications Using Ensemble Technique

Abstract

Abstract

The telecommunications industry understands the hyperlink between consumer pleasure and revenue. Customer attrition is a term used in the telecommunications industry to describe when a customercompany relationship ends. The churn forecast has recently intrigued stakeholders in the telecommunications industry. They learned that retaining customers is much more cost-effective than acquiring new ones. This paper presents an ensemble-based telecom churn predictive model of machine Learning (ML) algorithms such as XGBoost (extreme gradient boost), LightGBM, Random Forest (RF), and CatBoost. Using ML-based models for predictive analytics is very important in the telecommunications industry when it comes to predicting customer attrition. The ensemble method model helps the telephone company predict if a customer is likely to cancel. In addition, several algorithms such as XGBoost, LightGBM, RF, and Cat Boost have been integrated into this study using an ensemble technique called stacking, and metaheuristics have been developed with the telecoms business, to forecast customer attrition. According to the results of this study, the proposed method confirms performance in predicting customer churn with an accuracy of 92.2%.