Fulbright Scholar Derek Doran: Methods and Measurements for Complex Systems Security
Open talk for everyone
Where and when: Wednesday January 22 at 12:00 in room M326
Abstract
The ability for formal analysis and verification techniques to establish system security guarantees is inversely proportional to the inherent complexity of a system and the security and formal-method savviness of system designers and developers. It has further become evident to industry that the commonplace “patch the problem” approach to maintain security in live cyber systems is inadequate. These facts motivate an increasingly popular cybersecurity paradigm: do not seek total security, but accept that a system is insecure, and collect evidence from my system to inform investments minimizing a probability of attack or value of loss. This strategic approach demands sophisticated means to probe for data and characterize a cyber system, decision support models informing strategy that can be fit to data, and the ability for an analyst to rationalize cybersecurity recommendations from data.
I will highlight some of my recent methodological work along each of these demands including: (i) a largest-ever (as of 2019) probing and content-centric network analysis of the Tor dark web; (ii) a novel statistical network model for dynamic complex systems driven by seasonal processes; (iii) efforts in exploiting the topology of signals from deep learning systems as intuitive ‘explanations' of its decision making.
Bio
Derek Doran is an Associate Professor of Computer Science and Engineering at Wright State University and Director of Research at Tenet3 Cybernetics. His research interests are in developing methods that support the study and security evaluation of complex web and cyber-systems. His research has been supported by the U.S. National Science Foundation, Air Force Research Labs, Department of Energy, and the Ohio Federal Research Network. Derek is author of over 75 publications, including a text on network role mining and analysis, serves on the editorial board for Social Network Analysis and Mining, and is a founding chair of the NeSy special interest group on Explainable AI. He further manages the direction and supports the research scientists and hardware engineers of Tenet3, a well-funded and quickly growing cybersecurity firm working for the U.S. Dept of Defense and private sector customers.
See more at:
