Bounded Policy Iteration for Decentralized POMDPs.

Daniel S. Bernstein Eric A. Hansen Shlomo Zilberstein

Published in: IJCAI (2005)

Keyphrases

policy iteration
markov decision processes
dec pomdps
infinite horizon
reinforcement learning
policy iteration algorithm
markov decision problems
optimal policy
partially observable markov decision processes
finite state

dynamic programming
average reward
sample path
markov decision process
model free
distributed constraint optimization
partially observable
long run
state space
policy evaluation

fixed point
optimal control
least squares
multi agent
function approximation
continuous state
average cost
lost sales
temporal difference

reinforcement learning algorithms
reward function
search algorithm
bayesian networks
finite number
actor critic
np hard
dynamical systems
machine learning