A Machine Learning-Based Framework for Detecting Credit Card Anomalies and Fraud

Expeditious growth in e-commerce has emerged in an increasing number of online shopping societies. These shoppers rely on credit cards as a payment method or use a mobile wallet for their purchases. Thus, credit cards have become the main payment method in the e-world. Given the many transactions that occur daily, criminals propose tremendous opportunities to find different ways of attacking and stealing credit card information. Fraudulent credit card transactions are a serious business issue. These types of scams can result in significant economic and personal losses. As a result, businesses are increasingly devoted to developing new ideas and methods for detecting and preventing fraud, as well as acquiring their customer’s trust and protecting their privacy. In recent years, learning algorithms have emerged as important in research fields aimed at developing solutions to this issue. In this paper, the proposed framework will develop a two-stage anomaly and fraud detection model that uses a strengthened anomaly clustering algorithm in the first stage to find fraud patterns in transactions. The second stage uses a learning classifier to detect fraud in transactions based on fraud patterns and clients’ spending behavior. The transactions will be well balanced using hybrid sampling techniques to enhance the performance of the model. This framework also aims to optimize the learning classifier model, which will improve other existing models in terms of performance, computation time, and accuracy.