Iterated risk measures for risk-sensitive Markov decision processes with discounted cost.
Takayuki OsogamiPublished in: UAI (2011)
Keyphrases
- markov decision processes
- risk sensitive
- risk measures
- average cost
- optimal policy
- reinforcement learning
- finite state
- state space
- finite horizon
- policy iteration
- reinforcement learning algorithms
- dynamic programming
- infinite horizon
- average reward
- planning under uncertainty
- risk averse
- markov decision process
- utility function
- partially observable
- initial state
- action space
- multistage
- expected cost
- markov decision problems
- optimal control
- least squares
- data mining