Risk-sensitive average optimality in Markov decision processes.
Karel SladkýPublished in: Kybernetika (2018)
Keyphrases
- risk sensitive
- markov decision processes
- average cost
- finite state
- optimal policy
- finite horizon
- state space
- dynamic programming
- policy iteration
- stationary policies
- average reward
- optimal control
- reinforcement learning
- long run
- markov decision problems
- partially observable
- markov decision process
- initial state
- action space
- reward function
- decision processes
- reinforcement learning algorithms
- finite number
- control policy
- planning under uncertainty
- decision making
- optimality criterion
- linear programming