Experimental Design for Partially Observed Markov Decision Processes.
Leifur ThorbergssonGiles HookerPublished in: SIAM/ASA J. Uncertain. Quantification (2018)
Keyphrases
- experimental design
- markov decision processes
- partially observed
- state space
- finite state
- optimal policy
- policy iteration
- reinforcement learning
- empirical studies
- active learning
- dynamic programming
- finite horizon
- transition matrices
- sample size
- average cost
- markov decision process
- infinite horizon
- action space
- virtual learning environments
- average reward
- partially observable
- cooperative
- machine learning
- decision theoretic planning
- action sets