Andrew Clark, assistant professor of electrical and computer engineering, has received a three-year, $243,908 grant from the University of Washington and the Office of Naval Research for a research project titled “L2RAVE: Feedback-Driven Learn to Reason in Adversarial Environments for Autonomic Cyber Systems.”
The goal of Clark’s research is to use machine learning and artificial intelligence (AI) to create an autonomous security system that will better and more quickly identify cyber threats. This is an escalating problem for individuals trying to protect their identities and finances, as well as for companies protecting their competitive edge, and the government and military trying to protect their secrets and citizens.
Clark, who is taking on a new research focus by using AI in cybersecurity, is working to build an automated system, since it is difficult for humans to manage the size and complexity of cybersystems and the increasingly intelligent and persistent nature of malicious adversaries. By combining machine learning, data analysis, and AI, he will develop a framework that unifies statistical inferences with reasoning, so the system can make decisions with speed and flexibility.
The research is largely focused on government and Department of Defense network security, but Clark is also looking at how his work may apply to cyberphysical systems, such as autonomous vehicles. Security for driverless and connected vehicles would combine formal rules with learned information and real-time data. The research team includes a PhD student.
Award Date: July 30, 2018