Investigation of financial fraud detection by using computational intelligence

Abstract

Due to increasing technical capabilities, financial fraud becomes more sophisticated and more difficult to detect. As there are various categories and typologies of financial fraud, different detection techniques may be applied. However, based on the data generated daily by financial organizations, a technical solution must be implemented. This paper presents a comprehensive literature review of financial fraud, the categorizations of financial fraud, and financial fraud detection with the particular focus on computational intelligence-based techniques. As outlined in the reviewed literature, money laundering is a multilayered crime involving several fraud typologies; therefore, it was selected to be analysed in this research. The purpose of the research is to investigate the synthetic dataset of the money laundering scheme to see whether additional patterns could be outlined, which would help financial organizations to recognize suspicious activity easier. To achieve this goal, computational intelligence - decision tree, was selected as a classification method to identify additional patterns. As a result, data classification provides new data parameters which are essential in improving accurate and efficient financial fraud detection.