Q-ensemble learning for customer churn prediction with blockchain-enabled data transparency

Customer churn prediction is important for businesses, especially in the telecommunications sector, where retaining customers is more cost-effective than acquiring new ones. Traditional ensemble learning methods have enhanced prediction accuracy by combining multiple models, but they often struggle with efficiently processing complex, high-dimensional data. This paper introduces Q-Ensemble Learning, a novel approach integrating Quantum Computing (QC) with ensemble techniques to improve predictive performance. Our framework incorporates quantum algorithms such as Quantum Support Vector Machine (Q SVM), Quantum k-Nearest Neighbors (Q k-NN), Quantum Decision Tree (QDT), and others into an ensemble, leveraging their superior computational capabilities. The predictions from each quantum classifier are aggregated using a consensus voting mechanism and recorded on a blockchain to ensure robust data transparency and security. Extensive experiments on publicly available telecom customer churn datasets demonstrate that Q-Ensemble Learning improves accuracy by 15%, precision by 12%, and recall by 10% compared to classical ensemble methods such as Random Forest and Gradient Boosting. These metrics highlight the framework’s effectiveness in reducing false positives and negatives, significantly enhancing the reliability of churn prediction. The blockchain integration further ensures that the data handling process is transparent and secure, building trust in the model’s decisions.