Risk-sensitive infinite-horizon discounted piecewise deterministic Markov decision processes.
Yonghui HuangZhaotong LianXianping GuoPublished in: Oper. Res. (2022)
Keyphrases
- risk sensitive
- markov decision processes
- infinite horizon
- optimal policy
- finite horizon
- average cost
- stationary policies
- state space
- dynamic programming
- reinforcement learning
- policy iteration
- markov decision process
- finite state
- partially observable
- decision processes
- optimal control
- planning under uncertainty
- dec pomdps
- reinforcement learning algorithms
- long run
- action space
- average reward
- real valued
- decision problems
- search algorithm
- learning algorithm