Co-evolving Adversaries for Online Robust Malware Detection

Didi, Sabre and Nitschke, Geoff (2026) Co-evolving Adversaries for Online Robust Malware Detection, Proceedings of Genetic and Evolutionary Computation Conference (GECCO 2026), 13-17 July 2026, San José, Costa Rica, 537-540, ACM Press.

Text 2026-co-evolving-adversaries-for-online-robust-malware-detection.pdf Download (535kB)

Abstract

Adversarial robustness remains a significant challenge in machinelearning malware detection. We propose a co-evolutionary adversarial training framework that integrates gradient-based adversarial attacks with evolutionary controllers to improve robustness. A hybrid Convolutional Neural Network (CNN-MLP) processes spatial and vector-based features, while evolved controllers adaptively guide Projected Gradient Descent (PGD) using gradient statistics, optimizing adversarial effectiveness while respecting feature constraints. This online, co-evolutionary process exposes the classifier to increasingly adaptive attacks. Experimental results demonstrate that the proposed approach enhances adversarial robustness relative to standard training while maintaining clean accuracy, exemplifying a co-evolutionary arms-race framework for realistic, feature-constrained malware detection.

Item Type:	Conference paper
Subjects:	Computing methodologies > Artificial intelligence
Date Deposited:	15 Oct 2026 07:02
Last Modified:	15 Oct 2026 07:02
URI:	https://pubs.cs.uct.ac.za/id/eprint/1822

Actions (login required)

View Item