Toward error-bounded algorithms for infinite-horizon DEC-POMDPs.

Jilles Steeve Dibangoye Abdel-Illah Mouaddib Brahim Chaib-draa

Published in: AAMAS (2011)

Keyphrases

infinite horizon
dec pomdps
theoretical justification
partially observable markov decision processes
learning algorithm
markov decision processes
optimal control
long run
policy iteration
dynamic programming
query processing
efficient computation
markov decision problems