What is it about?
Cyber defense exercises help us see how well organizations can defend against online threats. By doing these tests, we find new weak spots and create better ways to protect against them. In this paper, we introduce a simulation using 'reinforcement learning.' This is a type of computer program that learns by trial and error, getting 'rewards' for good decisions and 'penalties' for bad ones. We use reinforcement learning to have computer programs 'learn' to attack and defend on their own. We also boost the program's smarts by giving it expert advice on cyber-attack methods and how to stop them. We've put this new method into practice using the CyberBattleSim system.
Featured Image
Photo by GuerrillaBuzz on Unsplash
Why is it important?
Attackers are evolving new ways to subvert our computing systems; our systems need to anticipate these and learn how to defend against them.
Read the Original
This page is a summary of: Knowledge Guided Two-player Reinforcement Learning for Cyber Attacks and Defenses, December 2022, Institute of Electrical & Electronics Engineers (IEEE),
DOI: 10.1109/icmla55696.2022.00213.
You can read the full text:
Contributors
The following have contributed to this page
