Cloud-Edge Training Architecture for Sim-to-Real Deep Reinforcement Learning.
Hongpeng CaoMirco TheileFederico G. WyrwalMarco CaccamoPublished in: IROS (2022)
Keyphrases
- reinforcement learning
- management system
- cloud computing
- learning capabilities
- supervised learning
- real world
- model free
- edge information
- training algorithm
- virtual machine
- optimal control
- edge detector
- markov decision processes
- image processing
- real time
- optimal policy
- data center
- training examples
- evaluation function
- online learning
- training set
- training process
- support vector
- machine learning