Enhancing Credit Card Fraud Detection under Severe Class Imbalance using Cost-Sensitive Learning and Threshold Optimization
DOI:
https://doi.org/10.4108/eetismla.12078Keywords:
credit card fraud detection, class imbalance, cost-sensitive learning, threshold calibration, Neural Networks, XGBoostAbstract
INTRODUCTION: Credit card fraud detection remains challenging because fraudulent transactions are rare, fraud patterns evolve over time, and precision–recall trade-offs vary under different class prevalences. While synthetic oversampling methods such as SMOTE are widely used, they may alter the observed feature distribution and complicate deployment interpretation. OBJECTIVES: This study develops a cost-sensitive and decision-threshold-calibrated fraud detection framework that preserves the original data distribution under severe class imbalance. METHODS: The framework combines cost-sensitive XGBoost and a Multi-Layer Perceptron trained with weighted Binary Cross-Entropy and Focal Loss. Hyperparameters are tuned using Optuna within a leakage-conscious validation protocol, and class imbalance is handled through scaleaware weighting rather than synthetic resampling. Decision thresholds are selected using minority-class F1 and an illustrative amount-aware cost criterion. SHAP analysis and a chronological split of the 2013 dataset are used to examine transformed-feature auditability and near-future generalization. RESULTS: On the imbalanced 2013 dataset, the optimized XGBoost model improves test-set PR-AUC from 0.7809 to 0.8815 and minority-class F1 from 0.7919 to 0.8497, with false positives decreasing from 21 to 13 on a test partition containing 98 fraud cases. On the balanced 2023 dataset, ranking performance is near-saturated and threshold selection mainly reduces false alarms. Under the illustrative assumption CFP = 1, amount-aware thresholding yields a lower simulated cost than the F1-optimized threshold, but with substantially more false positives. Chronological validation on the 2013 dataset yields lower PR-AUC than random stratified evaluation. CONCLUSION: The results suggest that fraud detection under class imbalance benefits from separating ranking optimization from decision-threshold selection. However, cost-aware and temporal findings should be interpreted as benchmark-based, deployment-motivated analyses rather than direct evidence of operational deployment readiness.
Downloads
References
[1] K. E. Hachimi, G. Orhanou, and L. Ballihi, “Credit card fraud detection model based on optimized long short-term memory: Imbalanced class solution and overfitting reduction,” Expert Systems with Applications, p. 130871, 2025.
[2] M. RezvaniNejad and A. S. Yameqani, “Wdae-gan: A hybrid dual autoencoder and generative adversarial framework with wavelet denoising for credit card fraud detection,” Expert Systems with Applications, p. 130078, 2025.
[3] R. K. Gupta, A. Hassan, S. K. Majhi, N. Parveen, A. T. Zamani, R. Anitha, B. Ojha, A. K. Singh, and D. Muduli, “Enhanced framework for credit card fraud detection using robust feature selection and a stacking ensemble model approach,” Results in Engineering, vol. 26, p. 105084, 2025.
[4] G. Metzler, X. Badiche, B. Belkasmi, E. Fromont, A. Habrard, and M. Sebban, “Tree-based cost sensitive methods for fraud detection in imbalanced data,” in International Symposium on Intelligent Data Analysis, pp. 213–224, Springer, 2018.
[5] W. Liu, H. Fan, M. Xia, and M. Xia, “A focal-aware cost-sensitive boosted tree for imbalanced credit scoring,” Expert Systems with Applications, vol. 208, p. 118158, 2022.
[6] J. L. Leevy, J. M. Johnson, J. Hancock, and T. M. Khoshgoftaar, “Threshold optimization and random undersampling for imbalanced credit card data,” Journal of Big Data, vol. 10, no. 1, p. 58, 2023.
[7] R. Ango, R. K. Masih, C. K. K. Reddy, M. Shuaib, M. Singh, and S. Alam, “Fraud detection in banking using the kaggle credit card dataset and xgboost model,” in 2024 International Conference on IoT Based Control Networks and Intelligent Systems (ICICNIS), pp. 968–973, IEEE, 2024.
[8] H. Huang, B. Liu, X. Xue, J. Cao, and X. Chen, “Imbalanced credit card fraud detection data: A solution based on hybrid neural network and clustering-based undersampling technique,” Applied Soft Computing, vol. 154, p. 111368, 2024.
[9] M. Herland, R. A. Bauder, and T. M. Khoshgoftaar, “The effects of class rarity on the evaluation of supervised healthcare fraud detection models,” Journal of Big Data, vol. 6, no. 1, p. 21, 2019.
[10] R. A. Bauder, T. M. Khoshgoftaar, and T. Hasanin, “An empirical study on class rarity in big data,” in 2018 17th IEEE international conference on machine learning and applications (ICMLA), pp. 785–790, IEEE, 2018.
[11] J. S. Sundar, R. Abinaya, M. Venkatesh, Y. Suthari, H. Koganti, and S. K. R. Katta, “Machine learning algorithms for real-time fraud detection in digital banking transactions,” in 2025 3rd International Conference on Intelligent Cyber Physical Systems and Internet of Things (ICoICI), pp. 787–792, IEEE, 2025.
[12] C. Elkan, “The foundations of cost-sensitive learning,” in International joint conference on artificial intelligence, vol. 17, pp. 973–978, Lawrence Erlbaum Associates Ltd, 2001.
[13] S. Elshaar and S. Sadaoui, “Cost-sensitive semi-supervised classification for fraud applications,” in International Conference on Agents and Artificial Intelligence, pp. 173–187, Springer, 2020.
[14] L. Huang, A. Abrahams, and P. Ractham, “Enhanced financial fraud detection using cost-sensitive cascade forest with missing value imputation,” Intelligent Systems in Accounting, Finance and Management, vol. 29, no. 3, pp. 133–155, 2022.
[15] J. Davis and M. Goadrich, “The relationship between precision-recall and roc curves,” in Proceedings of the 23rd international conference on Machine learning, pp. 233–240, 2006.
[16] J. Singla et al., “Comparing roc curve based thresholding methods in online transactions fraud detection system using deep learning,” in 2021 international conference on computing, communication, and intelligent systems (ICCCIS), pp. 9–12, IEEE, 2021.
[17] L. A. Dalton, “Optimal roc-based classification and performance analysis under bayesian uncertainty models,” IEEE/ACM Transactions on Computational Biology and Bioinformatics, vol. 13, no. 4, pp. 719–729, 2015.
[18] C. Wang, S. Chai, and H. Zhu, “Opendrift: Online evolving fraud detection for open-category and concept-drift transactions,” in 2023 IEEE International Conference on Web Services (ICWS), pp. 605–614, IEEE, 2023.
[19] M. L. G. ULB, “Credit card fraud detection: Anonymized credit card transactions labeled as fraudulent or genuine.” Kaggle Data Repository, 2013. Accessed: 2026- 02-23.
[20] K. Shenoy, “Credit card fraud detection dataset 2023: A comprehensive dataset for credit card fraud detection.” Kaggle Data Repository, 2023. Accessed: 2026-02-23.
[21] S. M. Lundberg and S.-I. Lee, “A unified approach to interpreting model predictions,” in Advances in Neural Information Processing Systems, vol. 30, 2017.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Vu Ngoc Thanh Sang
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
This is an open access article distributed under the terms of the CC BY-NC-SA 4.0, which permits copying, redistributing, remixing, transformation, and building upon the material in any medium so long as the original work is properly cited.
