Title: The effectiveness of cost sensitive machine learning algorithms in classifying Zeus flows

Authors: Ahmad Azab

Addresses: Department of Computer Engineering, American University of Middle East, Kuwait

Abstract: Zeus botnet is regarded as one of the primary sources of financial losses for both individuals and organisations. Therefore, the identification of its command and control (C&C) network traffic has become an important research field. Although the literature provided machine learning and other solutions for C&C identification, they suffer from various drawbacks. In this paper, we provide a framework that bridges the gap in terms of the machine learning solution, by building a classifier to detect the untrained version of Zeus botnet C&C traffic. The framework showed efficacy in detecting a new version of Zeus botnet, by building the classifier on an older version, compared to the machine learning approach used in the current research.

Keywords: Zeus; network; security; machine learning; botnet.

DOI: 10.1504/IJICS.2022.122378

International Journal of Information and Computer Security, 2022 Vol.17 No.3/4, pp.332 - 350

Received: 15 Jan 2019
Accepted: 05 May 2019
Published online: 22 Apr 2022 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article