Temporally Extended Metrics for Markov Decision Processes.
Philip AmortilaMarc G. BellemarePrakash PanangadenDoina PrecupPublished in: SafeAI@AAAI (2019)
Keyphrases
- decision theoretic planning
- markov decision processes
- temporally extended
- dynamic programming
- state space
- planning under uncertainty
- finite state
- optimal policy
- policy iteration
- finite horizon
- reinforcement learning algorithms
- reinforcement learning
- action space
- average cost
- partially observable
- decision processes
- infinite horizon
- markov decision problems
- probabilistic planning
- factored mdps
- orders of magnitude
- dec pomdps
- fixed point