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{210682,
        author = {Eshika Shukla and Sanchay Jangra and Durgesh Kumar and Dr. Bal Krishna Saraswat},
        title = {Fraud Detection in Credit Card Transactions Applying Machine Learning, including One more Comparative Study of Imbalance-Handling Techniques},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {6},
        pages = {1903-1912},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=210682},
        abstract = {Credit card fraud is typically rare but costly. Its rarity also makes it hard to detect - in the thoroughly applied ULB dataset, only 0.172% of transactions are dishonest, in actual application. The model that labels every transaction as genuine is therefore 99.8% correct and useless; this article is comparing three standard classifiers, Logistic Regression, Random Forest and XGBoost, each individual trained with and without SMOTE oversampling, under one fixed evaluation protocol, broadly speaking. In this broader context, precision, recall, F1, ROC-AUC and PR-AUC rather than precision are reported, employ SMOTE only to the training data so that no synthetic samples are reaching the test set, and apply SHAP to explain which characteristics drive the predictions; this goal is one fair, repeatable comparison that is revealing what imbalance handling actually changes, and at what cost in false alerts.},
        keywords = {credit card fraud detection, class imbalance, SMOTE, Random Forest, XGBoost, PR-AUC, SHAP, explainable machine learning},
        month = {November},
        }