Real-Time Data Stream Analytics and Artificial Intelligence for Enhanced Fraud Detection and Transaction Monitoring in Banking Security
American Journal of Innovation in Science and Engineering · 15.10.2025 · 4 (3) · 38-48
DOI: 10.54536/ajise.v4i3.5630 ↗Kurz gefasst
Fachartikel in der Zeitschrift American Journal of Innovation in Science and Engineering, online erschienen am 15.10.2025. Verfasst von Md Saiful Islam und 4 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref). Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Expanding rapidly as a result of the expansion in digital banking transactions, modern modes of financial fraud have grown more complex, and traditional rule-based systems of detecting fraud have proved inadequate because of high false-positive rates (normally 15-20 percent), slow response times (greater than 30 seconds), and unchanging detection signatures. It has been shown that this is a key weak point in all financial systems around the world, leading to multi-billion dollar losses every year, with the expeditious creation of dynamic detection frameworks that can occur in real-time being needed. We mitigated this basic challenge by developing and deploying the first end-to-end data analytics and ensemble AI models solution to produce streaming data analytics applications entirely optimized to detect fraud in high-velocity transaction spaces (produces >3,000 transactions/second). We tested seven machine learning architectures including new temporal convolutional network (TCN) and gradient-boosted LSTM hybrids by conducting extreme experimentation with synthetic (PaySim) and real-world (n=2.1 million transaction records) data. The optimized system demonstrated record performance levels: 98.7 percent in detection accuracy (p
Zitieren
APA
Islam, M. S., Ahmad, M. Y., Zerine, I., Biswas, Y. A., & Islam, M. M. (2025). Real-Time Data Stream Analytics and Artificial Intelligence for Enhanced Fraud Detection and Transaction Monitoring in Banking Security. American Journal of Innovation in Science and Engineering, 4(3), 38-48. https://doi.org/10.54536/ajise.v4i3.5630
Vancouver
Islam MS, Ahmad MY, Zerine I, Biswas YA, Islam MM. Real-Time Data Stream Analytics and Artificial Intelligence for Enhanced Fraud Detection and Transaction Monitoring in Banking Security. American Journal of Innovation in Science and Engineering. 2025;4(3):38-48. doi:10.54536/ajise.v4i3.5630
BibTeX
@article{Islam2025,
author = {Islam, Md Saiful and Ahmad, Md Yousuf and Zerine, Ismoth and Biswas, Younis Ali and Islam, Md Mainul},
title = {Real-Time Data Stream Analytics and Artificial Intelligence for Enhanced Fraud Detection and Transaction Monitoring in Banking Security},
journal = {American Journal of Innovation in Science and Engineering},
year = {2025},
volume = {4},
number = {3},
pages = {38--48},
doi = {10.54536/ajise.v4i3.5630}
}
EUVIMED