Influence-Based Policy Abstraction for Weakly-Coupled Dec-POMDPs.
Stefan J. WitwickiEdmund H. DurfeePublished in: ICAPS (2010)
Keyphrases
- dec pomdps
- decision theoretic planning
- infinite horizon
- markov decision problems
- partially observable markov decision processes
- optimal policy
- dynamic programming
- continuous state
- multi agent
- linear programming
- decision theoretic
- finite horizon
- theoretical justification
- single agent
- state dependent
- reinforcement learning
- partially observable
- planning under uncertainty
- efficient computation
- optimal control
- state space
- average cost
- long run
- finite state
- control policies
- action selection
- planning problems
- decision problems
- heuristic search
- optimal plans
- np hard