COST REDUCTION OF CREDIT CARD FRAUD DETECTION: USING INFORMATION FUSION APPROACH

Articles in Press

Document Type : Research Paper

Authors

1 University of Tehran

2 Faculty of Management, Tehran University, Tehran, Iran

3 4. MSc., Faculty of Management, Tehran University, Tehran, Iran

10.22059/frj.2024.338715.1007300

Abstract

Today, most companies and organizations have used e-commerce to gain productivity in their services and products in areas such as credit card, telecommunications, health insurance, car insurance and so on. Besides, due to the growing volume of credit card transactions, the demand for detecting fraud in this area is also increasing. The purpose of this paper is to provide a way to increase the accuracy of credit card fraud detection. The study was based on a dataset of transactions by a Brazilian bank over a two-month period from July 14th, 2004 to September 12th of the same year. In this paper, we obtain the cost function using the neural network and the K-Means clustering algorithm . According to the various indicators of fraud detection that have been introduced in the literature, we have selected the cost function introduced by Gadi, and have measured that. Since the use of only one algorithm has a high cost function, in order to reduce it, two kind of fusion has been implemented: Dempster-Shafer Evidence theory and probabilistic fusion. According to the results of the algorithms, the probabilistic fusion has been significantly reduced the cost function compared to Dempster-Shafer Evidence theory. This research has shown that fusion algorithms will have a much greater cost reduction than each of the algorithms alone.

Keywords

Main Subjects

Financial Research Journal

Articles in Press, Accepted Manuscript
Available Online from 04 January 2025

Files

Share

How to cite

Statistics