An Assessment on Credit Card Fraud Detection: Survey
Abhishek Malviya, Himanshu Yadav · 2020
Credit card fraud is a costly problem for many financial institutions, costing businesses billions of dollars a year. Many adversaries still escape fraud detection systems because these systems often do not include information about the adversary's knowledge of the fraud detection mechanism. This thesis aims to include information on the motivations of crooks and the knowledge base in an adaptive fraud detection system. In this thesis, we use a theoretical adversarial learning approach to classification to model the best fraudster strategy. We proactively adapt the fraud detection system to classify these future fraudulent transactions better. Therefore, this document aims to provide an over-supervised bird's-eye approach with a suitable feature extraction technique that improves fraud detection rather than mistakenly classifying an actual transaction as fraud.