Finite-Horizon Markov Decision Processes with State Constraints.
Mahmoud El ChamieBehçet AçikmesePublished in: CoRR (2015)
Keyphrases
- markov decision processes
- finite horizon
- state space
- optimal policy
- infinite horizon
- markov decision process
- optimal stopping
- action space
- reinforcement learning
- finite state
- partially observable
- average cost
- policy iteration
- dynamic programming
- decision theoretic planning
- average reward
- state abstraction
- transition matrices
- state dependent
- control policies
- real time dynamic programming
- state variables
- discounted reward
- single item
- decision problems
- multistage
- sufficient conditions
- search space