[talks] Talk today 3/12 - 4:15PM - Computer Science Room 105 -

Ginny Hogan gch at CS.Princeton.EDU
Wed Mar 12 09:09:24 EDT 2008


Reinventing Partially Observable Reinforcement Learning 
Eyal Amir 
University of Illinois, Urbana-Champaign 

Many complex domains offer limited information about their exact state and
the way actions affect them. There, autonomous agents need to make decisions
at the same time that they learn action models and track the state of the
domain. This combined problem can be represented within the framework of
reinforcement learning in POMDPs, and is known to be computationally
difficult. 
In this presentation I will describe a new framework for such decision
making, learning, and tracking. This framework applies results that we
achieved about updating logical formulas (belief states) after deterministic
actions. It includes algorithms that represent and update the set of
possible action models and world states compactly and tractably. It makes a
decision with this set, and updates the set after taking the chosen action.
Most importantly, and somewhat surprisingly, the number of actions that our
framework takes to achieve a goal is bounded polynomially by the length of
an optimal plan in a fully observable, fully known domain, under lax
conditions. Finally, our framework leads to a new stochastic-filtering
approach that has better accuracy than previous techniques. 

* Joint work with Allen Chang, Hannaneh Hajishirzi, Stuart Russell, Dafna
Shahaf, and Afsaneh Shirazi
(IJCAI'03,IJCAI'05,AAAI'06,ICAPS'06,IJCAI'07,AAAI'07). 




More information about the talks mailing list