HomeJournalsTBFLIVol. 1, Iss. 2Fraud Transaction Detection using Machine Learning
tbfli
Research ArticleTransactions on Banking, Finance, and Leadership Informatics

Volume 1, Issue 2 · 25 October 2025

ISSN: 3067-5804 · E-ISSN: 3067-5812

Fraud Transaction Detection using Machine Learning on Financial Datasets

Article ID:tbfli_25008

Abstract

Financial fraud poses a significant threat to the digital economy, with credit card fraud being a prevalent challenge. This study evaluates the performance of Logistic Regression (LR) and Extreme Gradient Boosting (XG Boost) models in detecting fraudulent transactions using 25 Oct 2025 (Published Online) financial datasets. The study uses practical data from 284,807 transactions, but only 492 are Fraud Detection, Machine Learning, Technique (SMOTE). Our findings show that XG Boost with Random Search selection is better XGBoost, Logistic Regression, and than Logistic Regression in all aspects. XG Boost yielded an accuracy of 99.96%, precision of Imbalanced Dataset (SMOTE) 95.11%, recall of 79.61%, and F1 score of 86.61%, while for Logistic Regression, the corresponding percentages were 99.92%, 88.1%, 60.5%, and 71.7%. The AUC statistic of 0.98 for XG Boost against 0.97 for LR classified the model as having better discriminant power. The results show that XG Boost is more suitable for real-time fraud detection. However, computational limitations and explainability issues should be considered. For future work, it is suggested that semi-supervised and supervised learning approaches be investigated and work with larger datasets to improve fraud detection in financial systems.

Keywords

Fraud DetectionMachine LearningXGBoostLogistic RegressionImbalanced Dataset (SMOTE)
View Full Article

Article Information

Received

9 September 2025

Accepted

16 October 2025

Published

25 October 2025

ISSN

3067-5804

E-ISSN

3067-5812

Article Type

Research Article

Open Access

Yes – Open Access