Enforcing Almost-Sure Reachability in POMDPs.
Sebastian JungesNils JansenSanjit A. SeshiaPublished in: CAV (2) (2021)
Keyphrases
- state space
- belief state
- partially observable markov decision processes
- reinforcement learning
- partially observable
- markov decision processes
- dynamic programming
- transitive closure
- continuous state
- point based value iteration
- belief space
- optimal policy
- global consistency
- distributed constraint optimization
- finite state
- markov decision problems
- dec pomdps
- predictive state representations
- data sets
- decision problems
- database systems
- databases
- policy gradient