Credit Card Risk Detection based on Feature-Filter and Fraud Identification

Abstract

Credit card fraud can destabilise economies, reduce confidence between customers and banks and affect other individuals or companies negatively. The primordial objective of banks is to identify fraudulent transactions with a high level of accuracy to reduce the training time and the costs of the manual investigation activity. This paper proposes a credit card fraud detection method using Random Forest as dimensionality reduction algorithm and Isolation Forest as a fraud detection algorithm. The method is applied to a large dataset in purpose to investigate their fraud detection accuracy. The experimental results of this study confirms the advantages and effectiveness of the proposed method in different criteria: accuracy, sensitivity and training time.