A Novel Deep Learning-Based Identification of Credit Card Frauds in Banks for Cyber Security Applications

Abstract

Due to the widespread use of constantly evolving internet technology and the increased frequency of cyber-attacks and crimes, cyber security is crucial for the banking sector. One of the biggest dangers confronting the banking sector globally is credit card (CC) fraud. It is becoming a serious issue and is growing rapidly, especially as the number of financial transactions utilizing CC keeps rising. The prevalence and growth of Internet banking have enhanced CC fraud identification. Finding fraudulent transactions of CC has become a major issue for internet buyers. In this study, an entirely novel deep learning (DL) algorithm is suggested for use in cyber security applications to identify CC thefts in the banking industry. We use a collection of significantly skewed CC fraud data sets to apply the proposed Multi-Gradient Whale Optimized Convolutional Neural Network (MW-CNN). The efficacy of the suggested methodis assessed depending on the performance evaluation criteria and comparing it with traditional techniques.