Hybrid ML Classification Approach for Customer Churn Prediction in Telecom Industry

Abstract

The rate at which customers abandon a product or service is known as the churn rate. The telecom industry must identify churn-risk customers in order to retain existing customers and maintain a higher competitive advantage. One of the main problems and major concerns for big companies is customer churn. Companies seeking to develop means to predict future customer churn especially in the telecom business because it directly affects their revenues. Customer lifetime value is greatly affected by churn rate because it affects the both company's future revenue and the length of service. Many companies have used different methods to predict churn rates and allow the development of effective customer retention plans because it cost much higher to get a new client rather than to retain an existing one. In order to predict customer churn in the telecom industry a Hybrid ML classification strategy is described in this research. Machine learning techniques are used to create the model in this paper. Customers who are likely to cancel their subscription can be predicted using machine learning algorithms. To predict churn, a combination of Support Vector Machine (SVM) and Naive Bayes (NB) algorithms is utilised. The main contribution of this work is the development of a churn prediction model that helps telecom operators to identify customers who are most likely to experience churn and better churn accuracy of the prediction.