Monday, March 4, 2013 - 4:00pm to 4:50pm
KEC 1001

Speaker Information

Matthew Taylor
Assistant Professor
Computer Science
Washington State University


Physical (robotic) agents and virtual (software) agents are becoming increasingly common in industry, education, and domestic environments. Recent research advances allow these agents can learn to complete tasks without human intervention. However, little is known about how humans should best teach such agents, nor how an agent could teach other agents. This unduly limits the rate at which the agents learn and reduces the potential benefits of leveraging existing human or agent knowledge.

This talk discusses some recent progress in enabling one agent to teach another reinforcement learning agent, even if the they have different learning methods and/or representations. Additionally, an approach that allows a human to teach an agent will be discussed, with an emphasis on limiting the amount of human effort required.

Speaker Bio

Matthew E. Taylor graduated magna cum laude with a double major in computer science and physics from Amherst College in 2001. After working for two years as a software developer, he began his Ph.D. work at the University of Texas at Austin with an MCD fellowship from the College of Natural Sciences. He received his doctorate from the Department of Computer Sciences in the summer of 2008, supervised by Peter Stone. Matt then completed a two year postdoctoral research position at the University of Southern California with Milind Tambe and spent 2.5 years as an assistant professor at Lafayette College in the computer science department. He is currently an assistant professor at Washington State University in the School of Electrical Engineering and Computer Science and is a recipient of the National Science Foundation CAREER award. Current research interests include intelligent agents, multi-agent systems, reinforcement learning, and transfer learning.