Finite-state approximations to constrained Markov decision processes with Borel spaces.
Naci SaldiSerdar YükselTamás LinderPublished in: Allerton (2015)
Keyphrases
- markov decision processes
- finite state
- policy evaluation
- optimal policy
- state space
- policy iteration
- dynamic programming
- reinforcement learning
- transition matrices
- markov decision process
- decision theoretic planning
- average cost
- infinite horizon
- action space
- reinforcement learning algorithms
- action sets
- planning under uncertainty
- average reward
- vector quantizer
- partially observable
- decision processes
- approximation methods
- stationary policies
- risk sensitive
- linear programming
- continuous state
- objective function
- learning algorithm