Anti-Money Laundering Machine Learning Pipelines; A Technical Analysis on Identifying High-risk Bank Clients with Supervised Learning

Anti - money laundering (AML) actions and measurements are among the priorities of financial institutions, for which machine learning (ML) has shown to have a high potential. In this paper, we propose a comprehensive and systematic approach for developing ML pipelines to identify high - risk bank clients in a dataset curated for Task 1 of the University of Toro nto 2023 - 2024 Institute for Management and Innovation (IMI) Big Data and Artificial Intelligence Competition. The dataset included 195,789 customer IDs, and we employed a 16 - step design and statistical analysis to ensure the final pipeline was robust. We also framed the data in a SQLite database, developed SQL - based feature engineering algorithms, connected our pre - trained model to the database, and made i t inference - ready, and provided explainable artificial intelligence (XAI) modules to derive feature importance. Our pipeline achieved a mean area under the receiver operating characteristic curve (AUROC) of 0.961 with a standard deviation (SD) of 0.005. Th e proposed pipeline achieved second place in the competition. Introduction In the contemporary financial landscape, money laundering represents a formidable challenge, compelling both financial institutions and regulatory bodies to seek innovative solutions. The integration of machine learning (ML) into anti - money laundering (AML) efforts has emerged as a promising avenue to enhance the detection and prevention of illicit financial activities. This paper investigates the technical considerations in employing supervised learning techniques to accurately identify high - risk bank clie nts, a critical component in the battle against money laundering. The utilization of ML for detecting money laundering transactions has shown significant promise. Jullum et al. developed an ML model that outperforms traditional systems by prioritizing transactions for manual investigation, using historic data from Norway ' s largest bank, DNB [1] .