Formal Controller Synthesis for Continuous-Space MDPs via Model-Free Reinforcement Learning.
Abolfazl LavaeiFabio SomenziSadegh SoudjaniAshutosh TrivediMajid ZamaniPublished in: ICCPS (2020)
Keyphrases
- continuous space
- controller synthesis
- model free reinforcement learning
- reinforcement learning
- markov decision processes
- discrete space
- multi agent
- mathematical morphology
- state space
- closed loop
- control system
- partially observable
- optimal policy
- control algorithm
- np hard
- dynamic programming
- gray level
- neural network
- learning tasks
- markov decision process
- learning algorithm