Credit card fraud detection using RUS and MRN algorithms

Abstract

Currently, enterprise systems have been focusing on expenditure services through credit card broadly because it is convenient and quick to pay for products and services. Thus, this research emphasizes on the fraud detection of credit card payment by using the machine learning technique called RUSMRN. The proposed method adopts three base classifiers which are MLP, NB and Naive Bayes algorithms. In addition, it can analyze the correctness to work with the unbalance datasets. Therefore, this research is focusing on the information of the credit card company of Taiwan for collecting data of customer behaviors in credit card payment. After that, it has brought the information to make prediction for correctness whether it has the risks in payment. The result shows that the proposed method can achieve the best classification performance in terms of accuracy and sensitivity.