Anytime Planning for Decentralized POMDPs using Expectation Maximization.
Akshat KumarShlomo ZilbersteinPublished in: UAI (2010)
Keyphrases
- expectation maximization
- partially observable markov decision processes
- dec pomdps
- em algorithm
- distributed constraint optimization
- belief space
- belief state
- stochastic domains
- partially observable
- sequential decision making problems
- predictive state representations
- multi agent
- optimal plans
- dynamic programming
- mixture model
- finite state
- reinforcement learning
- decentralized decision making
- probabilistic model
- planning problems
- point based value iteration
- decision theoretic
- cooperative
- maximum likelihood
- optimal policy
- decision theoretic planning
- markov decision problems
- ai planning
- planning under uncertainty
- continuous state
- partially observable markov decision process
- peer to peer
- motion planning
- action selection
- optimal planning
- initial state
- gaussian mixture