TOYOTA AI LAB SEMINAR
      Department of Electrical Engineering and Computer Science
              Stained-Glass Conference Room (3725 CSE)
              4:00 - 5:30 PM Tuesday, September 26, 2006


      Autonomous Bidding Agents: Strategies and Lessons from the
                     Trading Agent Competition **

                                  by
				   
                            Michael Wellman
	 Professor, Dept. of Computer Science and Engineering


Abstract:

Years of observing, analyzing, and competing in the Trading Agent
Competition (TAC) have taught us much about the potential of automated
bidders, and principles underlying their design.  I will share some of
the highlights of our experience with the TAC travel game, including:

1. a prevailing architecture based on prediction plus optimization,

2. a taxonomy of heuristics for bidding under uncertainty in
   interdependent markets, and

3. an empirical game-theoretic methodology for evaluating strategy
   ideas.

**This title and much of the content of the talk come from a book by
  this name written with Amy Greenwald and Peter Stone, forthcoming
  from MIT Press.