Logistics Regression Modelling for Anti Money Laundering Detection for Reporting Institution

Abstract

Money laundering is a significant threat to global financial systems and national security, and Malaysia's Anti-Money Laundering, Anti-Terrorism Financing and Proceeds of Unlawful Activities Act 2001 (AMLA 2001) mandates reporting institutions, such as banks and casinos, to identify and report suspicious transactions. Traditional rule-based approaches to Anti-Money Laundering (AML) detection face limitations, including high false-positive rates and a lack of adaptability to evolving criminal tactics. This research aims to address these challenges by identifying key parameters within the AMLA 2001 framework, developing a detection model that incorporates these parameters and leverages logistics regression techniques, and evaluating its accuracy, precision, recall, and capacity to reduce false positives. Employing a multi-pronged methodology, including literature review, data collection, model development, and evaluation, the study uncovered critical AMLA 2001 parameters predictive of illicit activities and created a model that enhances detection efficiency and accuracy. The findings have practical implications for any reporting institutions and regulatory bodies, contributing significantly to efforts to combat money laundering and safeguard the integrity of country's financial system.