Home // International Journal On Advances in Security, volume 14, numbers 1 and 2, 2021 // View article

Fraud Detection Using Multilayer Perceptron and Convolutional Neural Network

Authors:
Omoyele Odeniyi
Oghenerukvwe Oyinloye
Aderonke Thompson

Keywords: fraud; credit cards; Multi-layer perceptron; 1D- Convolutional Neural Network, Big Data.

Abstract:
In recent time, precarious transaction activities have attained a systematic daily occurrence, imbuing landed, personal and intangible properties. Of all these, credit card fraud is the most catastrophic if not detected on time for easy retrieval from the perpetrator. So, the threat actor gains unauthorized access in order to obtain money. Machine learning and data science has revolutionized and enhanced prompt discovery of expedient hidden information in data. Therefore, in this study, we develop an efficient fraud detection framework using non-rule-based approach of Multi-layer perceptron (MLP) on a given financial transaction dataset. Frauds were correctly predicted and detected. The algorithms on the datasets evaluates its effectiveness vis-à-vis frauds detection in bank transactions. The results are compared and evaluated using various evaluation metrics. In addition, we explored a 1D-Convolutional Neural Network, leveraging on its strength of less computational resource requirement. Observation from the experimental result revealed a desired gradual high accuracy.

Pages: 1 to 11

Copyright: Copyright (c) to authors, 2021. Used with permission.

Publication date: December 31, 2021

Published in: journal

ISSN: 1942-2636