Search |  Contact |  SRI Home Do not follow this link, or your host will be blocked from this site. This is a spider trap. Do not follow this link, or your host will be blocked from this site. This is a spider trap. Do not follow this link, or your host will be blocked from this site. This is a spider trap.A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A ASRI International.  333 Ravenswood Avenue.  Menlo Park, CA 94025-3493. SRI International is a nonprofit corporation.

AIC Seminar Series

Strategy Learning and Network Adaptation in Multi-Agent Systems

Marie desJardinsUniversity of Maryland, Baltimore County[Home Page]

Notice:  hosted by Sugato Basu

Date:  2006-07-06 at 11:00

Location:  EJ228  (Directions)

   Abstract

Networked multi-agent systems are comprised of many autonomous yet interdependent agents situated in a virtual social network. Examples of such systems include supply chains and sensor networks. In these systems, agents have a select set of other agents with whom they interact based on environmental knowledge, cognitive capabilities, resource limitations, and communications constraints. Previous findings have demonstrated that the structure of the artificial social network governing the agent interactions has a significant impact on the behavior and effectiveness of the agents. To improve individual and organizational performance, agents belonging to such communities can use two forms of organizational learning: they can adjust their strategies, given their position in the social network; and they can modify the social network by adding or removing connections to other agents. In this talk, I present our research on organizational learning for team formation in networked multi-agent systems. First, I describe agent-organized networks, a paradigm for locally adjusting network structure to improve system performance. Second, I describe a policy learning framework that agents can use to decide which teams to join, given their current position in the network. Both of these methods are shown to yield a significant increase in organizational team performance, compared to non-adaptive techniques. (joint work with Blazej Bulka and Matthew E. Gaston)

   Bio for Marie desJardins

Dr. Marie desJardins is an assistant professor in the Department of Computer Science and Electrical Engineering at the University of Maryland, Baltimore County. Her research is in artificial intelligence, focusing on the areas of machine learning, planning and scheduling, multi-agent systems, information management, reasoning with uncertainty, and decision theory.

Dr. desJardins received her Ph.D. in 1992 from the University of California, Berkeley. Her thesis research focused on developing techniques to allow autonomous agents to learn useful models of their environments without being supervised by a human designer. From 1991 to 2001, Dr. desJardins was a member of the research staff at SRI International. Her research there focused primarily on autonomous and mixed-initiative methods for planning, learning, and multi-agent systems. Dr. desJardins joined the UMBC faculty in August 2001. Current research projects include interactive and knowledge-intensive learning techniques, organizational learning and team formation in multi-agent systems, multiattribute optimization and visualization for school redistricting, and interactive graph layout.

Dr. desJardins has published over 60 scientific papers in journals, conferences, and workshops. She is the past Vice-Chair of ACMÂ’s SIGART, a former AAAI Councillor, and a member of the editorial boards of the Journal of Artificial Intelligence Research and AI Magazine.

Dr. desJardins can be contacted at the Dept. of Computer Science and Electrical Engineering, University of Maryland Baltimore County, 1000 Hilltop Circle, Baltimore MD 21250, mariedj@cs.umbc.edu, (410) 455-3967.

   On-line Resources