Customer churn prediction system: a machine learning approach

The customer churn prediction (CCP) is one of the challenging problems in the telecom industry. With the advancement in the field of machine learning and artificial intelligence, the possibilities to predict customer churn has increased significantly. Our proposed methodology, consists of six phases. In the first two phases, data pre-processing and feature analysis is performed. In the third phase, feature selection is taken into consideration using gravitational search algorithm. Next, the data has been split into two parts train and test set in the ratio of 80% and 20% respectively. In the prediction process, most popular predictive models have been applied, namely, logistic regression, naive bayes, support vector machine, random forest, decision trees, etc. on train set as well as boosting and ensemble techniques are applied to see the effect on accuracy of models. In addition, K-fold cross validation has been used over train set for hyperparameter tuning and to prevent overfitting of models. Finally, the obtained results on test set have been evaluated using confusion matrix and AUC curve. It was found that Adaboost and XGboost Classifier gives the highest accuracy of 81.71% and 80.8% respectively. The highest AUC score of 84%, is achieved by both Adaboost and XGBoost Classifiers which outperforms over others.

Automated summary

Customer churn prediction is a challenging issue in the telecom industry, and leveraging machine learning and artificial intelligence can significantly improve prediction accuracy. A proposed methodology with data preprocessing, feature analysis, model training, and evaluation through confusion matrix and AUC curve showed that Adaboost and XGboost classifiers performed the best.