GraphMDP: A New Decomposition Tool for Solving Markov Decision Processes.
Pierre LarochePublished in: Int. J. Artif. Intell. Tools (2001)
Keyphrases
- markov decision processes
- transition matrices
- semi markov decision processes
- state space
- optimal policy
- finite state
- average reward
- dynamic programming
- reinforcement learning
- reachability analysis
- stochastic shortest path
- infinite horizon
- factored mdps
- average cost
- policy iteration
- action space
- decision processes
- planning under uncertainty
- policy iteration algorithm
- finite horizon
- decision theoretic planning
- risk sensitive
- markov decision problems
- markov decision process
- reinforcement learning algorithms
- action sets
- partially observable
- reward function
- constraint satisfaction problems
- optimal solution