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{197810,
        author = {YUG SAINI and PRABHAT PRAJAPATI and HARSH RAMLANI and PRATIK BHOSALE},
        title = {A Review of ML techniques for Credit Card Fraud Detection},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {11},
        pages = {7064-7070},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=197810},
        abstract = {The recent surge in the adoption of digital payment systems has resulted in a substantial increase in the risk of credit card fraud faced by financial institutions. As the volume of transactions increases, the traditional rule-based monitoring systems have proven to be inadequate in identifying sophisticated patterns of fraud. Machine learning approaches have emerged as a practical solution for modelling complex transaction behaviour. The most important problem in dealing with fraud detection is the extreme imbalance in the proportion of normal to fraudulent transactions, which is less than 0.2%. As a result, the accuracy of the model is high, but the detection accuracy of fraudulent transactions is low. This paper critically examines recent studies on machine learning-based approaches for credit card fraud detection. Emphasis is given to the methods used for addressing the problems of class imbalance. Some of the widely used techniques such as Logistic Regression, Random Forest, and XGBoost have been analysed for their performance, interpretability, and computability. In addition, the performance of balancing techniques such as the Synthetic Minority Oversampling Technique (SMOTE) is also assessed based on the improvements reported in terms of Recall, F1-Score, and Precision-Recall AUC metrics for a set of benchmark problems. Apart from the performance metrics, the review also touches upon the practical aspects of the deployment of the models in a real-world scenario. The overall findings that can be obtained from the evidence indicate that the ensemble-based methods, in combination with the resampling techniques, can offer a better balance between the detection of fraud and false positives. The paper will conclude with the challenges that need to be addressed in the implementation of AI-based fraud detection methods.},
        keywords = {Credit Card Fraud Detection, Machine Learning, Class Imbalance, SMOTE, XGBoost, Ensemble Learning, Financial Security},
        month = {April},
        }