Interpretable credit card fraud detection using deep learning leveraging XAI

Abstract

Due to the internet's widespread accessibility, more and more businesses are bringing their offerings online. Besides, because of the growth of E-commerce websites, both individuals and businesses that deal in finances are more dependent on internet administrations to handle their business. Since more and more people are using online banking and making purchases online, credit card fraud has increased. Fraudsters can also use anything to disrupt the existing fraud detection system's systematic operation. As a result, we took on the issue of improving the existing fraud detection system to the highest possible level. This research seeks to develop an efficient fraud detection system by utilizing deep learning (DL) as well as the machine learning methods that are responsive to shifting patterns of customer behavior and have a tendency to reduce fraud manipulation through the identification and filtering of fraudulent activity in real time. The techniques in our research include Artificial Neural Network, Convolutional Neural Network, Recurrent Neural Network, Logistic Regression, K-Nearest Neighbor, Naive Bayes, Meta-Learning, and Explainable Artificial Intelligence (XAI). This research suggests that the K-Nearest Neighbor is the most effective algorithm with an accuracy of 99.75% among many others.