A dual layer explainable artificial intelligence approach for prediction and action in telecom customer churn analytics

Churn is a critical problem in the telecommunications industry with high requirements on the accuracy of the machine learning models used to address this issue. In this paper, we propose a framework that incorporates machine learning and explainable artificial intelligence (XAI) to solve the problem of customer churn in the telecommunications industry. Using the Telco Customer Churn dataset, we present an XGBoost model that has been optimized to achieve high accuracy in predicting the churn of customers in the telecommunications industry. Furthermore, SHAP and LIME are utilized to provide explanations for the model’s predicted outcomes for individual customers. The level of agreement between these two explanation methods is calculated using Spearman’s correlation coefficient and the Jaccard index. Based on the explanations provided by the XAI methods, the Strategic Actionability Score (SAS) is calculated to determine the business value of the model’s predictions. The results of this framework illustrate the potential of using XAI methods in the telecommunications industry to provide explanations for the predictions of machine learning models and to translate these explanations into actionable business insights. In addition, the proposed explainable AI-based churn prediction framework supports sustainable customer relationship management, efficient resource utilization, and long-term operational sustainability in the telecommunications industry by enabling telecom providers to optimize retention strategies and reduce unnecessary operational expenditure. Using this framework, telecommunications companies can develop responsible and transparent strategies to retain their customers.