Machine Learning Approaches for Cybersecurity Incident Measurement

February 4, 2021 - 12:30 PM to 1:30 PM
Zoom Meeting
Presented by: 
Dr. Benjamin J. Radford

The proliferation of advanced threat actors in cyberspace has left cybersecurity practitioners, engineers, and social scientists struggling to stay ahead of a phenomenon that is reshaping political conflict. The timely detection and attribution of these incidents is necessary to understand the threat environment, the actors therein, and the capabilities and objectives of those actors. Better attribution, along with more precise measurement of cybersecurity events in general, will not only lead to better legal and policy outcomes but also to an understanding of how political conflict occurs in a cyber-connected world. In this talk, Dr. Radford will discuss his research that applies machine learning to assist analysts with incident discovery and response. The talk will also explore the state of cybersecurity data available for quantitative social science research, the importance of having such data, and the challenges of acquiring it.


Benjamin J. Radford studies political conflict, cybersecurity, and the application of machine learning to problems in these domains. Dr. Radford has worked on several research programs for the United States Government including projects at DARPA and the Office of Naval Research. He was the principal investigator for a government-funded cybersecurity attribution program. His research has been published in the "Journal of Conflict Resolution" and "Political Science Research and Methods", among other venues. Dr. Radford received his PhD in political science from Duke University. He is currently an assistant professor of Political Science and Public Administration at UNC Charlotte.