AI Seminar

Finding Preferable Choices in Constrained and Multiagent Contexts

Ed Durfee

Computer Science and Engineering, University of Michigan
Tuesday, November 18, 2008
4:00pm - 5:30pm
Stained-Glass Conference Room (3725 Beyster Bldg.)

Add to Google Calendar

About the Event

Motivated by problems that arise in coordinating people with cognitive difficulties, we have been investigating techniques for finding individual and group choices that are (arguably) most preferable given constraints. In particular, we have been studying possibilities and limitations in finding preferred outcomes based only on qualitative information about preferences, captured in CP-nets. Further, there can be constraints on allowable combinations of features of outcomes, which introduces further complications to the search process. In this talk, I will summarize the motivations and assumptions behind our efforts, describe the underlying representations upon which our work is based, and discuss the algorithms that we have developed, along with our evaluations of their performance, highlighting lessons learned and remaining open problems.

Additional Information

Sponsor(s): Toyota AI Seminar Series

Open to: Public