In the world of financial transactions, rule based heuristics are often employed to detect fraud, rather than to detect anomalies. Given the fundamental shortcomings of rule based heuristics, how does machine learning help in fraud detection?

Bayesian inferencing enables calculation of the probability of a new event on the basis of earlier probability estimates of events in the past, derived from existing empiric data. With the Bayesian approach, one can use objective data or subjective opinion to specify a response. Once this is done, the earlier probabilities can then be used to make better decisions. AI-powered AML solutions thus derive meaning from complex or imprecise data, recognise patterns, and determine trends that most human beings or other computer techniques fail to identify. In addition, they become increasingly intelligent as more and more data enters the system, thus rendering them far more effective than conventional AML solutions in the long run.