Empirical Analysis of Machine Learning Algorithms on Detection of Fraudulent Electronic Fund Transfer Transactions

Abstract

Online financial systems like electronic funds transfers (EFT) are heavily dependent on the public Internet. A diverse set of Internet traffic is generated from the use of different online financial platforms. The analysis of raw Internet traffic fails to identify any anomalous behavior in financial transactions. A similar result is obtained when we apply application layer anomaly detection in financial transactions. In this article, we propose a machine learning-based multi-layer framework which will detect and classify anomalous financial transactions. The proposed framework can help a financial service provider to avoid incidents like intrusions and online frauds. It also provides a secure mechanism to detect network anomalies in financial transactions to augment the credibility of such online financial platforms or gateways.