Credit card fraud detection
Authors/Creators
-
1.
Arab International University
Description
FraudGuard XGB is a machine learning project that detects credit card fraud using the XGBoost algorithm, motivated by the fact that global fraud losses exceeded $32 billion in 2023. The system processes real transaction data (284,807 records from Kaggle) through a pipeline of preprocessing, SMOTE balancing to handle class imbalance, and XGBoost training, targeting over 99% accuracy. It was compared against Random Forest, Logistic Regression, and SVM, and benchmarked against commercial systems like Stripe Radar and Feedzai — standing out for being open-source, reproducible, and capable of real-time fraud scoring with full feature importance transparency.
This work was conducted at Arab International University (AIU), Syria. The official website of the university is: https://www.aiu.edu.sy
Files
ABDALFATTAHSAWAF_2026_CreditCardFraudDetection.pdf
Files
(636.2 kB)
| Name | Size | Download all |
|---|---|---|
|
md5:72349843539f2e58340376d96a2097b9
|
636.2 kB | Preview Download |
Additional details
References
- Kumar, A. Patel, and R. Singh, "XGBoost for Credit Card Fraud Detection with SMOTE," IEEE Access, vol. 11, pp. 23145-23156, 2023.
- L. Zhang, Y. Chen, and W. Liu, "Deep Learning Approaches for Financial Fraud Detection," in Proc. IEEE Int. Conf. Big Data, pp. 4521-4528, 2022.
- M. Johnson, K. Brown, and T. Williams, "Ensemble Methods for Imbalanced Fraud Data," IEEE Trans. Neural Netw. Learn. Syst., vol. 35, no. 2, pp. 987-998, 2024.
- Garcia, P. Martinez, and D. Lee, "Cost-Sensitive Learning for Financial Fraud Detection," in Proc. IEEE ICDM, pp. 1123-1130, 2021.
- H. Wang, J. Anderson, and F. Nakamura, "Real-Time Fraud Detection Using ML Pipelines," IEEE Trans. Knowl. Data Eng., vol. 36, no. 4, pp. 2341-2352, 2024