Point-based incremental pruning heuristic for solving finite-horizon DEC-POMDPs.
Jilles Steeve DibangoyeAbdel-Illah MouaddibBrahim Chaib-draaPublished in: AAMAS (1) (2009)
Keyphrases
- infinite horizon
- finite horizon
- dec pomdps
- dynamic programming
- optimal policy
- markov decision problems
- markov decision processes
- long run
- partially observable markov decision processes
- combinatorial optimization
- optimal control
- markov decision process
- partially observable
- average cost
- state space
- reinforcement learning
- lead time
- finite state
- heuristic search
- admissible heuristics
- search algorithm
- decision making under uncertainty
- decision problems
- steady state
- approximate solutions
- multistage
- markov chain
- special case
- lower bound
- control policies
- optimal solution