Artificial Intelligence based Fraud Agent to Identify Supply Chain Irregularities

A.J. Hoffman and R.E. Tessendorf (South Africa)

Keywords

supply chain, fraud agent, neural networks, decision trees

Abstract

: Risk Management is of increasing importance in global supply chains; while up to 10% of goods are lost between manufacturing and customers, the markets of legal players are cannibalised by greyand black-market products. The primary weakness in existing systems is the inability to provide accurate and reliable fraud indicators when irregular behaviour is suspected. This paper demonstrates a Supply Chain Fraud Agent that not only detects the presence of abnormalities, but also provides accurate identification of the colluding parties in irregular behaviour. It is shown that the simplistic interpretation of discrepancy indicators can lead to an unacceptably large number of false alarms. The paper then demonstrates how the application of artificial intelligence techniques can reduce the number of false alarms to such an extent that effective action can be taken.

Important Links:



Go Back