The seminars generally will be held every Friday at noon in Room 105, College of Engineering West Hall on the Monroe Park Campus. For more information, please contact our seminar coordinator - Tarynn Witten, MSW, M.D., Ph.D.
Use Adobe Acrobat Reader to download the seminar fliers and collaterals.
If you need assistance accessing any of the information on this page, please contact the Marketing & Communications department.
Explainable AI (XAI) and Goal Reasoning
Friday, Nov. 30, 2018 | Noon–1 p.m. | Engineering East Hall, Room E1232
Speaker: Dr. David W. Aha
Abstract: Goal reasoning allows an agent to dynamically reason about the relative utilities of goals it can pursue, which may result in changing its active goals. Its applications include autonomous vehicle control, agent control in simulations and game AI, and intelligent decision aids. The goal reasoning process, which broadly supports deliberative autonomy, is enpowering but requires (in addition to AI safety constraints) the agent to explain its reasoning. This relates to the popular topic of Explainable AI (XAI). In this talk, I’ll summarize the objectives and status of DARPA’s XAI program and then review some of our group’s work on goal reasoning, focusing on situation assessment and decision making tasks. I’ll finish by briefly discussing our initial steps on explainable goal reasoning.
Bio: David W. Aha serves as Acting Director of NRL’s Navy Center for Applied Research in AI, as well as its section on Adaptive Systems. His research interests include intelligent (e.g., goal reasoning) agents, symbolic planning, case-based reasoning, explainable AI (XAI), machine learning, and related topics. He has co-organized 37 international events related to these topics (e.g., AAAI-19 Senior Member Track, IJCAI-18 XAI Workshop), launched the UCI Repository for ML Databases, served as an AAAI Councilor, co-created the AAAI Video Competition, received awards for five publications, and gave the IAAI-17 Engelmore Memorial Lecture. He has led 4 DARPA (e.g., XAI) or ONR evaluation teams.