Technical - Partner Event
AI Seminar – Zaheen Farraz Ahmad
When
April 9, 2021 - April 9, 2021
12:00 PM – 1:00 PM MST
Where
Online
Online
Abstract: Sample-based planning is a powerful family of algorithms for generating intelligent behavior from a model of the environment. Generating good candidate actions is key to the success of sample-based planners, especially in large action spaces. Candidate actions usually i) exhaust the entire search space, ii) are handcrafted using domain knowledge or, iii) are produced using a learned policy. We show how we can explicitly learn a candidate action generator by optimizing a novel objective, marginal utility. The marginal utility measures the increase in utility of an action over previous actions. We validate our candidate action generator in domains with both continuous and discrete action spaces. We show how planning with an explicit candidate generator outperforms planners instantiated with handcrafted actions, trained stochastic policies, and other natural objectives for generating actions.
Presenter Bio: Zaheen is a PhD student at the University of Alberta working with Michael Bowling and Levi Lelis. His research interests lie in artificial intelligence with a stronger focus on search, learning and games. His current work is exploring how search and learning methods can be effectively combined to produce stronger agents for adversarial domains. Zaheen also received his MSc at the University of Alberta working with Robert Holte and he is the lead graduate student member of the Computer Curling Research Group.
The University of Alberta Artificial Intelligence (AI) Seminar is a weekly meeting where researchers (including students, developers, and professors) interested in AI can share their current research. Presenters include local speakers from the University of Alberta and industry as well as other institutions. The seminars discuss a wide range of topics related in any way to Artificial Intelligence, from foundational theoretical work to innovative applications of AI techniques to new fields and problems are of interest. Learn more at the AI Seminar website and by subscribing to the mailing list!
Looking to build AI capacity? Need a speaker at your event?
