Structured-policy Q-learning: an LMI-based Design Strategy for Distributed Reinforcement Learning.
Lorenzo SforniAndrea CamisaGiuseppe NotarstefanoPublished in: CDC (2022)
Keyphrases
- reinforcement learning
- optimal policy
- multi agent
- function approximation
- action selection
- state space
- cooperative
- reinforcement learning algorithms
- learning algorithm
- optimal control
- markov decision processes
- design process
- peer to peer
- distributed systems
- control policy
- reward function
- partially observable
- policy iteration
- state action
- reinforcement learning methods
- actor critic
- policy search