Optimizing Memory-Bounded Controllers for Decentralized POMDPs

Christopher Amato Daniel S. Bernstein Shlomo Zilberstein

Published in: CoRR (2012)

Keyphrases

reinforcement learning
distributed constraint optimization
dec pomdps
multi agent
peer to peer
control system
dynamic programming

partially observable markov decision processes
memory requirements
limited memory
cooperative
optimal policy
markov decision processes
bounded memory

memory size
state space
distributed systems
belief state
memory usage
main memory
computing power

partially observable
memory space
control law
decision making
continuous state
search algorithm
linear programming