EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Real-Time Data Stream Analytics and Artificial Intelligence for Enhanced Fraud Detection and Transaction Monitoring in Banking Security

Md Saiful Islam, Md Yousuf Ahmad, Ismoth Zerine, Younis Ali Biswas, Md Mainul Islam

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).

AbstractBibliografische AngabenZitierenVerwandte Studien

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}
}
Kostenlos starten →