Analyzing and Escaping Local Optima in Planning as Inference for Partially Observable Domains.
Pascal PoupartTobias LangMarc ToussaintPublished in: ECML/PKDD (2) (2011)
Keyphrases
- partially observable domains
- partially observable
- sensing actions
- reinforcement learning
- planning problems
- inverse reinforcement learning
- partially observable markov decision processes
- decision theoretic
- ai planning
- evolutionary algorithm
- optimal solution
- markov decision processes
- decision problems
- situation calculus
- event calculus
- action models
- graphical models
- heuristic search
- probabilistic reasoning
- bayesian networks