Application of Machine Learning Algorithms to Enhance Money Laundering and Financial Crime Detection
Loading...
Date
2011-09
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
The British University in Dubai (BUiD)
Abstract
The underlying idea of this thesis is to understand the current challenges and difficulties that financial institutions face with regards to the prevention and detection of financial crime and suspicious activities in relation to fraud, money laundering and terrorist financing. It sheds light upon contemporary developments in financial crime activities and the anti-money laundering regulations, policies and frameworks that have been set in order to address this issue and overcome the associated challenges and difficulties. The collection of information in relation to financial crime activities alongside adopted existing regulations would facilitate the identification of the weaknesses and flaws that constitute the areas for enhancement. The investigation process follows the scientific method approach and hence starts with a background (introduction) of financial crime history and its types including fraud, money laundering, market manipulation, insider trading which might be used to finance terrorist activities. The literature review would cover the details and particulars of each type of financial crime such as money laundering, fraud, market manipulation and insider training. It would also cover the current methodologies and technologies used to prevent and detect financial crime and suspicious financial activities. The background overview serves as pillars to support the research aim, objectives and questions. In order to analyze the performance of machine learning algorithms, data was provided by a bank to be used for educational purposes and shall remain undisclosed. The data was used as training and testing sets to analyze certain machine learning algorithms in terms of performance (cost / benefit analysis) and accuracy (mean error square and confusion matrix). The research is concluded with a conclusion section which recapitulates the results obtained and observations with regards to the current detection mechanisms and the applied machine learning algorithms. The recommendation section emphasizes the steps that can be taken and improvements to the existing methodologies and tools used in the prevention and detection of financial crimes.
Description
Keywords
machine learning algorithms, money laundering, financial crime detection,, lesson observation, United Arab Emirates (UAE), teacher performance, school leaders, public schools