Sensitivity-based nested partitions for solving finite-horizon Markov decision processes.
Weiwei ChenPublished in: Oper. Res. Lett. (2017)
Keyphrases
- finite horizon
- markov decision processes
- transition matrices
- optimal policy
- semi markov decision processes
- infinite horizon
- optimal stopping
- average cost
- finite state
- policy iteration
- reinforcement learning
- markov decision process
- state space
- dynamic programming
- decision theoretic planning
- markov decision problems
- partially observable
- decision problems
- reward function
- action space
- control policies
- multistage
- expected reward
- stochastic shortest path
- decision making
- linear programming
- search space