Scalable photonic reinforcement learning by time-division multiplexing of laser chaos.
Makoto NaruseTakatomo MihanaHirokazu HoriHayato SaigoKazuya OkamuraMikio HasegawaAtsushi UchidaPublished in: CoRR (2018)
Keyphrases
- reinforcement learning
- function approximation
- state space
- optimal policy
- markov decision processes
- learning algorithm
- reinforcement learning algorithms
- temporal difference
- neural network
- model free
- autonomous learning
- stochastic approximation
- video streams
- action selection
- optimal control
- highly scalable
- web scale
- reinforcement learning methods
- learning process
- multi agent
- physical characteristics
- robotic control