Exploration Policies for On-the-Fly Controller Synthesis: A Reinforcement Learning Approach.
Tomás DelgadoMarco Sánchez SorondoVíctor A. BrabermanSebastián UchitelPublished in: ICAPS (2023)
Keyphrases
- reinforcement learning
- controller synthesis
- optimal policy
- autonomic computing systems
- policy search
- active exploration
- multi agent
- markov decision process
- action selection
- state space
- function approximation
- partially observable markov decision processes
- control algorithm
- reward function
- fitted q iteration
- closed loop
- control policy
- markov decision problems
- markov decision processes
- learning algorithm
- reinforcement learning algorithms
- control system
- machine learning
- optimal control