Structure Learning in Ergodic Factored MDPs without Knowledge of the Transition Function's In-Degree.
Doran ChakrabortyPeter StonePublished in: ICML (2011)
Keyphrases
- structure learning
- factored mdps
- graphical models
- bayesian networks
- transition model
- knowledge representation
- parameter estimation
- basis functions
- transfer learning
- belief propagation
- approximate inference
- context specific
- knowledge base
- state space
- generative model
- markov networks
- conditional independence
- prior knowledge