The Complexity of Policy Evaluation for Finite-Horizon Partially-Observable Markov Decision Processes.
Martin MundhenkJudy GoldsmithEric AllenderPublished in: MFCS (1997)
Keyphrases
- partially observable markov decision processes
- policy evaluation
- finite horizon
- optimal policy
- markov decision processes
- decision problems
- infinite horizon
- policy iteration
- finite state
- reinforcement learning
- state space
- average cost
- multistage
- dynamic programming
- sufficient conditions
- long run
- markov decision process
- least squares
- partially observable
- monte carlo
- model free
- computational complexity
- initial state
- policy gradient
- action space
- multi agent
- control system
- special case
- belief state
- decision theoretic
- optimal control
- worst case