Correct-by-construction policies for POMDPs.
Nils JansenSebastian JungesJoost-Pieter KatoenTim QuatmannBernd BeckerRalf WimmerLeonore WintererPublished in: SNR (2019)
Keyphrases
- partially observable markov decision processes
- optimal policy
- reinforcement learning
- policy gradient methods
- predictive state representations
- dynamical systems
- planning under uncertainty
- decision problems
- policy search
- belief state
- decision making
- markov decision problems
- distributed constraint optimization
- markov decision processes
- dynamic programming
- partially observable
- construction process
- point based value iteration
- finite state
- markov decision process
- continuous state