Recursively-Constrained Partially Observable Markov Decision Processes.
Qi Heng HoTyler J. BeckerBenjamin KraskeZakariya LaouarMartin FeatherFederico RossiMorteza LahijanianZachary N. SunbergPublished in: CoRR (2023)
Keyphrases
- partially observable markov decision processes
- finite state
- belief state
- planning under uncertainty
- reinforcement learning
- belief space
- decision problems
- dynamical systems
- optimal policy
- continuous state
- markov decision processes
- dynamic programming
- partially observable stochastic games
- partial observability
- planning problems
- state space
- partially observable domains
- stochastic domains
- predictive state representations
- partially observable markov
- partially observable
- partially observable markov decision process
- general purpose
- point based value iteration
- dec pomdps
- initial state
- special case
- computational complexity
- multi agent