A Unified Framework for Risk-sensitive Markov Decision Processes with Finite State and Action Spaces
Yun ShenSteffen GrünewälderKlaus ObermayerPublished in: CoRR (2011)
Keyphrases
- state and action spaces
- markov decision processes
- risk sensitive
- action space
- state space
- optimal policy
- finite state
- reinforcement learning
- policy iteration
- average cost
- average reward
- reinforcement learning algorithms
- infinite horizon
- partially observable
- finite horizon
- dynamic programming
- markov decision problems
- decision processes
- planning under uncertainty
- reward function
- markov decision process
- learning algorithm
- optimality criterion
- search space