Abstraction and Refinement for Solving Continuous Markov Decision Processes.
Alberto ReyesPablo H. IbargüengoytiaLuis Enrique SucarEduardo F. MoralesPublished in: Probabilistic Graphical Models (2006)
Keyphrases
- markov decision processes
- decision theoretic planning
- transition matrices
- semi markov decision processes
- action space
- state abstraction
- markov decision problems
- state space
- optimal policy
- stochastic shortest path
- continuous state spaces
- planning under uncertainty
- policy iteration
- finite horizon
- partially observable
- reachability analysis
- finite state
- dynamic programming
- macro actions
- reinforcement learning
- average reward
- decision processes
- markov decision process
- reinforcement learning algorithms
- infinite horizon
- risk sensitive
- factored mdps
- average cost
- action sets
- state and action spaces
- model based reinforcement learning
- decision makers
- np hard