Credit Card Fraud Detection Using Quantum Machine Learning

Abstract

The rise of digital payment systems has increased the incidence of financial fraud. Conventional fraud detection systems, relying upon preset rules and parameters, may not be equipped to respond in the face of changing fraud patterns. We propose a hybrid fraud detection mechanism that utilizes various machine learning algorithms alongside other behavioral and geo-velocity checks. This model uses both SVM and QSVM hybrid to evaluate transaction risk real time. Depending on the score generated towards the chance of a transaction being fraudulent, they will either be approved, flagged for additional verification or blocked. Tools used for the implementation of this system is based on various Python frameworks and quantum computing. The experimental results prove that the proposed method achieves better performance in fraud detection with less false alarms than traditional methods.