Enhanced Telecom Customer Churn Prediction Using Machine Learning and Deep Learning Models

Customer churn prediction is the greatest priority for telecom companies wishing to retain their customers and minimize revenue loss. Thus, in this paper, several machine learning and deep learning models have been used for predicting the churn of telecom customers. These include the Random Forest, Logistic Regression, AdaBoost, Gradient Boosting, and Deep Neural Networks (DNN). In doing so, an emphasis was placed on feature selection, data preprocessing, and class balancing within the text, using the Kaggle Telecom Customer Churn dataset. The presented results highlight the accuracy enhancement abilities of ensemble learning and DNNs while performing experiments at different metrics such as accuracy, precision, recall, and F1-score. Furthermore, it also highlights the point of interpretability and understanding churning behavior in telecom settings by deploying predictive models in real-world telecom settings.