An Effective Classifier for Predicting Churn in Telecommunication

J. Pamina, J. Beschi Raja, S. Sathya Bama, S. Soundarya, M.S. Sruthi, S. Kiruthika, V.J. Aiswaryadevi and G. Priyanka

In recent days, telecom industry plays a major role in our daily life. The proliferation of telecommunication industry becomes very difficult for the service providers to survive in the market. To stabilize in this field, the service providers have to be aware of the features that make the customer to churn. The proposed predictive model identifies the traits that highly influence customer churn, with the help of machine learning techniques like KNN, Random Forest and XG Boost. IBM Watson dataset has been analysed to forecast the churn. At last a comparative study has been made among the machine learning algorithm to identify the better algorithm of higher accuracy. The proposed model shows that Fiber Optic customers with greater monthly charges have higher influence for churn.

Volume 11 | 01-Special Issue

Pages: 221-229