What do taxation, cyber networks, software and humans have in common? They are all vulnerable to adversarial conflicts. Taxation faces noncompliance, networks face attacks, software faces malware, and humans-- among many vulnerabilities --fall victim to disinformation. In this talk, Dr. Una-May O'Reilly will describe work on Artificial Adversarial Intelligence (AAI), employing machine learning and evolutionary algorithms to support the discovery of new knowledge about these systems' adversarial nature.
Una-May O'Reilly is the leader of the Anyscale Learning For All (ALFA) Group at the MIT Computer Science and Artificial Intelligence Lab. ALFA focuses on machine learning technology, evolutionary algorithms and data science for knowledge mining, prediction, analytics and optimization. ALFA also conducts research in cyber security and software analysis. Dr. O'Reilly is the author of over 150 academic papers and has served as vice chair of ACM Special Interest Group on Evolutionary Computation (ACM-SIGEVO) and chair of the largest international Evolutionary Computation Conference (GECCO). She serves as a senior editor for several major journals and as an advisor to national laboratories and startups. Her Ph.D. at Carleton University was one of the world’s first on the AI topic of genetic programming.