Using Bayesian Belief Networks for Credit Card Fraud Detection

L.E. Mukhanov (Russia)

Keywords

Prediction with expert advice, Bayesian Networks, Fraud detection.

Abstract

The number of credit card fraud cases is permanently increasing. Therefore, at this moment the problem of fraud prevention system implementation with automatic decision maker functions is really actual. In this paper the disadvantage of Bayesian Belief Networks for the fraud detection and the developed input data representation method are considered. At the end of this paper the results of the evaluative testing and corresponding conclusions are described.

Important Links:



Go Back