Credit Card Fraud Detection Based on Deep Neural Network Approach

Abstract

Currently, the financial companies’ reliance has become on the new technologies and developments, in which these companies provide the customers with various services, electronically. One the most common financial services is the credit cards, where the customers can complete any financial payments using a credit card, such as; Visa card and MasterCard. However, criminals started to find ways to attack customers’ cards for fraud purposes, and that has caused a big issue for both of companies and customers. This work aim to propose a new for credit card fraud detection based on a dataset for customers’ transactions of a well-known company called Vesta, where the utilization of deep learning approaches has achieved a method with efficient performance in identifying transactions that was classified as fraud with 99.1% score of area under ROC curve.