Fraud Detection in Banking Electronic Payment Using Data Mining

Fraud Detection in Banking Electronic Payment Using Data Mining

Marzie Hosseini1

1) IT Expert at Tourism Bank postgraduate from Amirkabir university

IOI Code : XYCC-ABHDG
Publication : International Conference on Science and Engineering(icesconf.com)
Abstract :
    Fraud continuing variation is one of the most important Fraud Detection challenge in Electronic Payment Systems. Electronic markets need high Performance Fraud Detection methods with top accuracy. In this project we have used logistic regression, BP neural network and GMDH neural network to make a useful method detecting Fraud in a bank financial transactions. We have implemented these methods on real dataset and have measured their results. The results are GMDH neural network 91.73%, logistic regression 89.26% and BP neural network 90.54%. Therefore our recommended method has the best accuracy.

Keywords : fraud electronic payment systems financial transactions data mining gmdh neural network