Title: Algorithmic Problems in Sequential Decision Making
Speaker: Michael Littman, Rutgers University
Date: November 8, 2004 3:30-4:30pm
Location: DIMACS Center, CoRE Bldg, Room 431, Rutgers University, Busch Campus, Piscataway, NJ
My research is concerned with making decisions to maximize utility. There have been many models that capture various attributes of sequential decision problems, and some interesting efficient algorithms are known. I'll survey a set of models and what is known about how well they can solved, highlighting a number of open algorithmic challenges related to Markov decision processes and game theory. I'm hoping that recent developments such as the quantum computation model and semidefinite programming might be able to shed light on these problems, so I welcome your input.