Home / Current Issue / Paper 1722685
A Comparative Evaluation of Machine Learning Models for Credit Card Fraud Detection in E-Commerce
Subject area: Science,Engineering and Technology · Area of research: Machine Learning
DOI: 10.64388/IREV10I2-1722685
Abstract
The rapid growth of e-commerce and electronic payment services has increased the convenience of digital transactions while also creating more opportunities for fraudulent activity. Conventional fraud detection approaches based on manual review and fixed rules can struggle with large transaction volumes, changing fraud patterns, and false-positive alerts. This study comparatively evaluates four supervised machine learning algorithms-Logistic Regression, K-Nearest Neighbors (KNN), Decision Tree, and Support Vector Machine (SVM) for credit card fraud detection. The study used a labelled transaction dataset and followed an iterative Agile-based development process involving data preparation, model training, testing, and performance evaluation. The models were assessed using accuracy, precision, recall, F1-score, receiver operating characteristic area under the curve (ROC-AUC), and training time. The reported experimental results show that Logistic Regression achieved the highest accuracy (63.33%), precision (67.61%), and ROC-AUC (0.6855), while KNN produced the highest recall (67.50%) and F1-score (0.6506). SVM produced moderate results but required the longest training time of 0.0345 seconds. Decision Tree had the shortest training time at 0.0058 seconds but recorded the weakest overall predictive performance, including 54.00% accuracy and 40.00% recall. The findings demonstrate that model selection for fraud detection should not be based on accuracy alone because missing fraudulent transactions can have serious financial and operational consequences. Logistic Regression provided the strongest overall balance among the evaluated models, although the moderate performance of all four algorithms indicates that larger and more representative datasets, improved feature engineering, class-imbalance treatment, and hyperparameter optimization are required before practical deployment.
Keywords
credit card fraud, e-commerce, machine learning, fraud detection, supervised learning
References
[1] Adepoju, O., Wosowei, J., Lawte, S., & Jaiman, H. (2019). Comparative evaluation of credit card fraud detection using machine learning techniques. In 2019 Global Conference for Advancement in Technology (GCAT) (pp. 1–6). IEEE. Crossref
[2] Bolton, R. J., & Hand, D. J. (2002). Statistical fraud detection: A review. Statistical Science, 17(3), 235–255. Crossref
[3] Bhattacharyya, S., Jha, S., Tharakunnel, K., & Westland, J. C. (2011). Data mining for credit card fraud: A comparative study. Decision Support Systems, 50(3), 602–613. Crossref
[4] Ngai, E. W. T., Hu, Y., Wong, Y. H., Chen, Y., & Sun, X. (2011). The application of data mining techniques in financial fraud detection: A classification framework and an academic review of literature. Decision Support Systems, 50(3), 559–569. Crossref
[5] Bahnsen, A. C., Aouada, D., & Ottersten, B. (2015). Example-dependent cost-sensitive decision trees. Expert Systems with Applications, 42(19), 6609–6619. Crossref
[6] Carcillo, F., Dal Pozzolo, A., Le Borgne, Y.-A., Caelen, O., Mazzer, Y., & Bontempi, G. (2018). SCARFF: A scalable framework for streaming credit card fraud detection with Spark. Information Fusion, 41, 182–194. Crossref
[7] Cortes, C., & Vapnik, V. (1995). Support-vector networks. Machine Learning, 20(3), 273–297. Crossref
[8] Cover, T., & Hart, P. (1967). Nearest neighbor pattern classification. IEEE Transactions on Information Theory, 13(1), 21–27. Crossref
[9] Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5–32. Crossref
[10] Friedman, J. H. (2001). Greedy function approximation: A gradient boosting machine. The Annals of Statistics, 29(5), 1189–1232. Crossref
[11] Chawla, N. V., Bowyer, K. W., Hall, L. O., & Kegelmeyer, W. P. (2002). SMOTE: Synthetic minority over-sampling technique. Journal of Artificial Intelligence Research, 16, 321–357. Crossref
[12] Dal Pozzolo, A., Caelen, O., Johnson, R. A., & Bontempi, G. (2015). Calibrating probability with undersampling for unbalanced classification. In 2015 IEEE Symposium Series on Computational Intelligence (pp. 159–166). IEEE. Crossref
[13] Carcillo, F., Le Borgne, Y.-A., Caelen, O., Kessaci, Y., Oblé, F., & Bontempi, G. (2021). Combining unsupervised and supervised learning in credit card fraud detection. Information Sciences, 557, 317–331. Crossref
[14] Uchechukwu Sophia, E. (2026). Fraud detection system using machine learning models in e-commerce (Unpublished master's thesis). Niger Delta University.
How to cite this paper
@article{1722685,
author = {Ziakegha Lucky Tonbrapagha, Ziakegha Clement Ebimobowei},
title = {A Comparative Evaluation of Machine Learning Models for Credit Card Fraud Detection in E-Commerce},
journal = {Iconic Research And Engineering Journals},
year = {2026},
volume = {10},
number = {2},
pages = {3745-3750},
issn = {2456-8880},
url = {https://www.irejournals.com/formatedpaper/1722685.pdf},
abstract = {The rapid growth of e-commerce and electronic payment services has increased the convenience of digital transactions while also creating more opportunities for fraudulent activity. Conventional fraud detection approaches based on manual review and fixed rules can struggle with large transaction volumes, changing fraud patterns, and false-positive alerts. This study comparatively evaluates four supervised machine learning algorithms-Logistic Regression, K-Nearest Neighbors (KNN), Decision Tree, and Support Vector Machine (SVM) for credit card fraud detection. The study used a labelled transaction dataset and followed an iterative Agile-based development process involving data preparation, model training, testing, and performance evaluation. The models were assessed using accuracy, precision, recall, F1-score, receiver operating characteristic area under the curve (ROC-AUC), and training time. The reported experimental results show that Logistic Regression achieved the highest accuracy (63.33%), precision (67.61%), and ROC-AUC (0.6855), while KNN produced the highest recall (67.50%) and F1-score (0.6506). SVM produced moderate results but required the longest training time of 0.0345 seconds. Decision Tree had the shortest training time at 0.0058 seconds but recorded the weakest overall predictive performance, including 54.00% accuracy and 40.00% recall. The findings demonstrate that model selection for fraud detection should not be based on accuracy alone because missing fraudulent transactions can have serious financial and operational consequences. Logistic Regression provided the strongest overall balance among the evaluated models, although the moderate performance of all four algorithms indicates that larger and more representative datasets, improved feature engineering, class-imbalance treatment, and hyperparameter optimization are required before practical deployment.},
keywords = {credit card fraud, e-commerce, machine learning, fraud detection, supervised learning},
month = {August},
doi = {https://doi.org/10.64388/IREV10I2-1722685}
}