Solving DEC-POMDPs by Expectation Maximization of Value Function.
Zhao SongXuejun LiaoLawrence CarinPublished in: AAAI Spring Symposia (2016)
Keyphrases
- expectation maximization
- dec pomdps
- em algorithm
- sequential decision making problems
- theoretical justification
- reinforcement learning
- dynamic programming
- probabilistic model
- maximum likelihood
- markov decision problems
- machine learning
- image segmentation
- linear programming
- sufficient conditions
- decision making under uncertainty