Generalized and Bounded Policy Iteration for Interactive POMDPs.
Ekhlas SonuPrashant DoshiPublished in: ISAIM (2012)
Keyphrases
- policy iteration
- markov decision processes
- reinforcement learning
- policy iteration algorithm
- optimal policy
- average reward
- finite state
- markov decision problems
- model free
- sample path
- partially observable markov decision processes
- markov decision process
- partially observable
- fixed point
- infinite horizon
- state space
- dynamic programming
- actor critic
- temporal difference
- continuous state
- least squares
- policy evaluation
- reinforcement learning algorithms
- policy search
- reward function
- action space
- optimal control
- decision problems
- average cost
- radial basis function
- search algorithm
- multi agent