142
click to vote
CDC
2010
IEEE
14 years 9 months ago
2010
IEEE
We give an optimal dynamic programming algorithm to solve a class of finite-horizon decentralized Markov decision processes (MDPs). We consider problems with a broadcast informati...