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{205611,
        author = {LAKHAN SINGH and laxman singh},
        title = {A Multi-Stage Hybrid Gradient Boosting Model with Optimal Selection Strategy for Online Fraud Detection},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {1},
        pages = {7728-7735},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=205611},
        abstract = {The increasing popularity of digital payment systems has led to a rise in online fraud, increasing the demand for accurate and scalable fraud detection solutions. Fraud detection remains challenging due to severe class imbalance between fraudulent and legitimate transactions, the continuously evolving nature of fraud patterns, and the requirement for transaction verification within milliseconds. A hybrid machine learning framework is proposed to detect fraudulent transactions by integrating XGBoost and CatBoost classifiers using a soft-voting ensemble strategy. Principal Component Analysis is applied to reduce input dimensionality while preserving 95% of the original data variance and enhancing model performance. XGBoost demonstrates strong learning capability on transformed numerical features, while CatBoost improves robustness through effective regularization and reduced overfitting. The hybrid framework combines probabilistic outputs from both classifiers to improve prediction confidence. Performance evaluation is conducted against individual boosting models using accuracy, precision, recall, F1-score, and ROC-AUC metrics. Experimental results indicate that the hybrid model achieves superior performance, including higher fraud recall and reduced false positives, demonstrating suitability for real-time online transaction fraud detection systems.},
        keywords = {Online payment fraud detection, machine learning, ensemble learning, XGBoost, CatBoost, Principal Component Analysis, class imbalance, real-time transaction analysis.},
        month = {June},
        }