Regret-Optimal Control under Partial Observability.
Joudi HajarOron SabagBabak HassibiPublished in: CoRR (2023)
Keyphrases
- optimal control
- partial observability
- reinforcement learning
- partially observable
- infinite horizon
- online learning
- lower bound
- dynamic programming
- reward function
- markov decision process
- belief space
- planning problems
- control strategy
- belief state
- optimal control problems
- partially observable markov decision processes
- learning agent
- markov decision processes
- state space
- real time
- neural network
- average cost
- machine learning
- partial information
- reinforcement learning algorithms
- solving problems
- planning domains
- dynamical systems
- orders of magnitude
- mobile robot
- markov decision problems
- objective function