Improving Credit Card Fraud Detection using a Meta-Classification Strategy

Abstract

One of the issues facing credit card fraud detection systems is that a significant percentage of transactions labeled as fraudulent are in fact legitimate. These " false alarms " delay the detection of fraudulent transactions and can cause unnecessary concerns for customers. In this study, over 1 million unique credit card transactions from 11 months of data from a large Canadian bank were analyzed. A meta-classifier model was applied to the transactions after being analyzed by the Bank's existing neural network based fraud detection algorithm. This meta-classifier model consists of 3 base classifiers constructed using the decision tree, naïve Bayesian, and k-nearest neighbour algorithms. The naïve Bayesian algorithm was also used as the meta-level algorithm to combine the base classifier predictions to produce the final classifier. Results from the research show that when a meta-classifier was deployed in series with the Bank's existing fraud detection algorithm improvements of up to 28% to their existing system can be achieved.