Your Cart

Your cart is empty

Home  >  Volume 31 (July 2015)

Credit Card Fraud Detection Using Firefly Algorithm by Olusegun Folorunso by (page 101 – 106)
Sale price: $5.00
Quantity:

Abstract

In this study, a Firefly Algorithm (FA) based Credit Card Fraud Detection (CCFD) for e-payment system is presented. This takes into consideration, discrete optimization problem associated with arbitrary changes in the state of credit card holder’s behaviour variables that occur as payment transactions progresses. However, the existing CCFD techniques always lead to false fraud alerts and eventual misclassification of transaction decisions.    Genetic Algorithm (GA), a bio-inspired algorithm and other computational intelligence techniques have been used earlier for CCFD. From literature, it has proven that FA is superior to GA because of its efficient global search strength. Hence, FA, a meta-heuristic algorithm is proposed in this study and to the best of our knowledge, FA has never been used in detecting credit card fraud.Meanwhile, FA takes inputs from account details of credit cards andfraud rules setto classify frauds based on Critical Value Identification (CVI) for each credit card. CVI determines light intensity and minimum attractiveness (brightest firefly) used to develop the FA objective function neededto minimize false fraud alert being experienced in CCFD system.

Keywords: Fraud detection system, electronic payment system, credit card, firefly algorithm, critical value identification.

Close
Loading...