Generating Trading Agent Strategies
Google engEDU
52 min – Jan 17, 2006

Google TechTalks
January 17, 2006

Daniel M. Reeves

Daniel Reeves recently completed his PhD in Computer Science at the University of Michigan as a student of Michael Wellman and is now (temporarily) a lecturer at Michigan, teaching Knowledge-Based Systems (Lisp, Prolog, and Mathematica for AI Programming). His most active area of research is the application of game-theoretic and computational techniques to strategic behavior in games, particularly for eCommerce-inspired market mechanisms. He is one of the creators of and top competitors in the international Trading Agent Competition. Dr Reeves is also one of the top ultra-marathon inline skaters in the US and climbs stairs competitively.

ABSTRACT
A Strategy Generation Engine is a system that reads a description of a game or market mechanism and outputs strategies for participants. Ideally, this means a game solver—an algorithm to compute Nash equilibria. This is a well-studied problem and very general solutions exist, but they can only be applied to small, finite games. I will present methods for finding or approximating Nash equilibria for infinite games, and for intractably large finite games.
video
http://video.google.com/videoplay?docid=5301380251515556722

• Download: “Generating Trading Agent Strategies” (video/mp4)

September 10, 2009 EDIT

Its rather interesting to study the co-notation of Dr Daniel Reeves’ game theory with Poker. Here is a list of expected value for poker hands I found on Google.