Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{189194,
        author = {Nikhil Kumar},
        title = {“An Explainable Hybrid Machine Learning Framework for Customer Churn Prediction”},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {7},
        pages = {5045-5052},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=189194},
        abstract = {Customer churn represents a significant challenge for organizations operating in highly competitive industries, particularly within the telecommunications and financial service sectors, where customer retention directly influences profitability and long-term sustainability. Timely identification of customers at risk of churn enables organizations to design proactive retention strategies and minimize revenue loss. Despite notable advancements in machine learning-based churn prediction, many existing approaches rely on black-box models that lack transparency, limiting their adoption in real-world business decision-making. To address this challenge, this study proposes an explainable hybrid machine learning framework for customer churn prediction that effectively balances predictive performance and interpretability.
The proposed framework employs a two-stage hybrid architecture. In the first stage, feature selection techniques are utilized to identify the most relevant attributes, thereby reducing data dimensionality and eliminating redundant or irrelevant features. In the second stage, an ensemble-based machine learning classifier is applied to perform churn prediction. To enhance model interpretability, an explainable artificial intelligence (XAI) layer based on Shapley Additive Explanations (SHAP) is integrated, enabling both global and local interpretation of feature contributions to churn predictions.
The effectiveness of the proposed framework is validated using a real-world customer churn dataset and benchmarked against traditional machine learning models. Experimental results demonstrate that the hybrid model consistently outperforms baseline approaches in terms of accuracy, F1-score, and ROC-AUC, while simultaneously providing actionable insights into the key factors influencing customer churn. The results indicate that integrating explainability within hybrid machine learning models not only improves predictive accuracy but also enhances transparency, trust, and usability for business stakeholders. This study underscores the practical value of explainable hybrid frameworks in customer analytics and offers a robust foundation for future research on interpretable predictive modeling.},
        keywords = {Customer Churn Prediction; Hybrid Machine Learning; Explainable Artificial Intelligence (XAI); Feature Selection; SHAP-Based Interpretability; Data Analytics.},
        month = {December},
        }