A Binary Firefly Optimized Stacking Ensemble Model for Customer Churn Prediction in the Telecommunications Industry

Predicting Telecom Customer Churn and Win‑Back Probability: A Stacked Ensemble Approach

XGBoost Algorithm for Telecom Customer Churn Prediction and Its Business Implications

Predicting customer loss is a critical challenge for telecommunications companies, as identifying at-risk customers is key to reducing financial losses and formulating effective retention strategies. Conventional statistical methods and basic ensemble models often perform poorly on telecom datasets due to their inability to capture complex feature relationships and handle class imbalance. This study applies the Extreme Gradient Boosting (XGBoost) algorithm to telecom customer churn prediction, comparing its performance with Logistic Regression and Random Forest using the public Telco Customer Churn dataset (7,043 records, 21 features). A rigorous data preprocessing pipeline was implemented, including missing value handling, categorical encoding, and feature standardization, with an 80/20 train-test split that preserved class distribution. Hyperparameter tuning for XGBoost addressed the 26.5% churn rate imbalance. Evaluation metrics (accuracy, precision, recall, F1- score, AUC) showed XGBoost outperformed benchmark models: 82.1% accuracy, 78.3% churn recall, 59.2% precision, 67.4% F1-score, and 0.869 AUC. Feature importance analysis identified customer tenure, contract type, monthly charges, and fiber optic internet service as the primary churn drivers. XGBoost also captured complex non-linear interactions unrecognized by other models.

Customer Churn Prediction in the Telecommunications Sector Using Explainable Artificial Intelligence

: Customer churn poses a significant threat to profitability in the saturated telecommunications industry, yet accurately predicting churn remains challenging due to high-dimensional data and class imbalance where churners represent a small minority. While complex ensemble methods achieve high accuracy, their ”black-box” nature limits business adoption, as practitioners require transparent insights to design effective retention campaigns. This paper proposes a comprehensive Machine Learning pipeline that bridges the gap between predictive performance and model interpretability. We implement and compare five classifiers—Logistic Regression, Random Forest, XGBoost, LightGBM, and CatBoost—optimized via Bayesian hyperparameter tuning and evaluated using the recall-focused F2-score to address class imbalance. Our results demonstrate that gradient boosting models, particularly XGBoost, outperform aggregation-based ensemble strategies, achieving the highest F2-score of 0.7500 with a recall of 0.9385. Crucially, we integrate SHAP-based Explainable Artificial Intelligence to provide both global and local interpretability, revealing that contract type, tenure, and technical support subscriptions are the primary churn drivers. This dual-layer transparency enables targeted retention strategies while preserving high recall in imbalanced telecom data.

Machine Learning–Based Customer Churn Prediction in Banking Using Feature Selection and Ensemble Models

Customer churn is a big challenge in the banking industry as there are certain strategies need to retain a customer forever. There are different computational analyses and reports to estimate possible churn and modify service tactics accordingly, resulting in greater customer retention rates. This research has proposed a novel Hilbert-Schmidt Independence Criterion (HSIC) amidst other techniques for the selection of the intricate features for a robust predictive performance. Alongside, this study is examined with customized machine learning models and evaluated with major key evaluation metrics to checkmate its optimal performance. The dataset utilized is collected from Kaggle comprises of 3,380 entries 19 predictor variables. The customized bagging approach incorporated with the HSIC optimized features performed excellently, achieving an accuracy of 98.06% and an AUC value of 0.9984 while reducing training time by 19% compared to the whole feature set. Ensemble techniques consistently beat single classifiers across all feature subsets, and dimensionality reduction considerably accelerated training while retaining an accuracy loss of less than 1%. Overall, this proposed method is effective at predicting customer attrition risk, delivering useful information for financial organizations when developing customer retention strategies, allowing banks to better personalize service approaches to keep clients.

AI-based Customer Churn Prediction System using XGBoost for Early Retention and Revenue Optimization

The study aims to develop an AI-based customer churn prediction system using the XGBoost algorithm to improve prediction accuracy and enable early identification of customers who are likely to leave a service. A total of 2000 customer records were used for the analysis. Two categories were considered for comparison; Group 1 employed conventional machine learning models such as Decision Tree and Logistic Regression, while Group 2 implemented the XG Boost algorithm on the same dataset to capture complex customer behavior patterns. Data preprocessing steps including data cleaning, encoding, feature scaling, and class imbalance handling were applied to both groups before model training and testing. Performance evaluation was carried out using accuracy, precision, recall, and F1-score metrics. The XGBoost model achieved superior results with an accuracy of 93.2%, precision of 92.6%, recall of 94.1%, and F1-score of 93.3 when compared to traditional machine learning methods. The results demonstrate that the proposed XGBoost-based system is highly effective in identifying high-risk churn customers at an early stage. Hence, the proposed model provides better predictive performance and supports improved decision-making for customer retention strategies and business growth.

An Intelligent Machine Learning Framework for Customer Churn Prediction in CRM Systems

In today's highly competitive market environment, organizations face significant challenges in retaining customers due to increasing competition, evolving customer expectations, and unpredictable churn behavior. In order to address these concerns and predict customer turnover using the Telco Customer turnover dataset, this article suggests a CRM system that relies on clever machine learning techniques. To ensure high-quality data for model training, the proposed system incorporates thorough data pretreatment stages such ADASYN class balancing, missing value management, label encoding, and z-score normalisation. Build and test two supervised ML models, XGBoost and Random Forest (RF), to see how well they perform. The trials showed that the Proposed RF Model achieved 97.1% accuracy (ACC), 97.6% precision (PRE), 97.4% recall (REC), and 97.2% F1-score (F1), whereas the Model Proposed XGBoost Model generated 96.9% ACC, 96.7% PRE, 96.6% REC, and 96.3% F1-score. The suggested method outperforms state-of-the-art machine learning and deep learning models for predicting customer attrition. With the help of the suggested framework, businesses may pinpoint consumers who are likely to churn, which in turn allows for more proactive retention measures, happier customers, and more profits in the long run.