Cloud-Edge Training Architecture for Sim-to-Real Deep Reinforcement Learning.
Hongpeng CaoMirco TheileFederico G. WyrwalMarco CaccamoPublished in: CoRR (2022)
Keyphrases
- reinforcement learning
- real time
- multiscale
- learning capabilities
- training process
- supervised learning
- cloud computing
- training algorithm
- management system
- software architecture
- learning algorithm
- feedforward artificial neural networks
- virtual machine
- function approximation
- optimal policy
- dynamic programming
- multi agent
- data center
- training samples
- online learning
- training set
- network architecture
- deep architectures
- image processing