Fighting Money Laundering with Statistics and Machine Learning

Syeda Nazia Banu, Shaik Abdul Anees, Chitikela Madhu Gangadhar, Kasarapu Rajeshwar Reddy, Nallagatla Vamshi, Boyini Avinash

2025

Abstract

Money laundering is a massive issue it’s when criminals take their dirty cash and try to make it look clean by shuffling it through what seem like everyday transactions. Every year, billions of dollars get laundered this way, creating a real mess for the global financial system. The usual way banks and regulators try to catch this involves setting up rules like flagging any transaction over $10,000. Trouble is, these rules aren’t all that clever. They end up pointing the finger at a ton of innocent transactions, which annoys customers and piles extra work on banks, in our research, we’ve come up with a fresh, smarter way to tackle this problem. We’ve built a system that mixes two big ideas: supervised learning, where we train a computer to spot money laundering by showing it examples of legit and shady transactions, and anomaly detection, which is all about catching stuff that doesn’t fit the normal flow like a huge payment suddenly heading to some offshore account. But we didn’t just leave it there (G. King and S. Lewis, 2020) (J. West and M. Bhattacharya, 2016). We threw in some slick statistical tricks, custom made for digging into financial data, to help our model get a better grip on how money moves (P. G. Campos and E. S. de Almeida, 2018) and how accounts are linked up. For example, our system keeps an eye on when transactions happen and how different accounts are tied together. If a bunch of accounts are tossing money around in a weird loop or some other odd pattern, that’s a signal something might be up, to see if this actually works, we tested it with fake transaction. Data and stuff, we cooked up to look like real money laundering setups. This let us play around without stepping on anyone’s privacy. The payoff? Our approach did a better job at nabbing the sketchy stuff and didn’t hassle nearly as many innocent folks as the old rule-based setups or even some other machine learning attempts. This project is part of a larger push to sharpen the tools banks and regulators use to fight money laundering. By making these systems brainier and more on-point, we’re helping put a dent in how criminals exploit the financial world, keeping things safer for everyone.

Paper Citation

in Harvard Style

Banu S., Anees S., Gangadhar C., Reddy K., Vamshi N. and Avinash B. (2025). Fighting Money Laundering with Statistics and Machine Learning. In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25; ISBN 978-989-758-777-1, SciTePress, pages 744-752. DOI: 10.5220/0013889400004919

in Bibtex Style

@conference{icrdicct`2525,
author={Syeda Banu and Shaik Anees and Chitikela Gangadhar and Kasarapu Reddy and Nallagatla Vamshi and Boyini Avinash},
title={Fighting Money Laundering with Statistics and Machine Learning},
booktitle={Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25},
year={2025},
pages={744-752},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013889400004919},
isbn={978-989-758-777-1},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25
TI - Fighting Money Laundering with Statistics and Machine Learning
SN - 978-989-758-777-1
AU - Banu S.
AU - Anees S.
AU - Gangadhar C.
AU - Reddy K.
AU - Vamshi N.
AU - Avinash B.
PY - 2025
SP - 744
EP - 752
DO - 10.5220/0013889400004919
PB - SciTePress