Dynamic Programming for POMDP with Jointly Discrete and Continuous State-Spaces.
Donghwan LeeNiao HeJianghai HuPublished in: ACC (2019)
Keyphrases
- continuous state spaces
- dynamic programming
- state space
- partially observable markov decision processes
- continuous state
- reinforcement learning
- markov decision processes
- dec pomdps
- optimal policy
- markov decision problems
- control problems
- heuristic search
- action space
- optimal control
- belief state
- partially observable
- finite state
- markov decision process
- dynamical systems
- infinite horizon
- planning problems
- planning under uncertainty
- markov chain
- robot navigation
- reinforcement learning algorithms
- reward function
- decision problems
- linear programming
- state dependent
- hidden state
- supervised learning
- neural network