Step-Ahead Error Feedback for Distributed Training with Compressed Gradient.
An XuZhouyuan HuoHeng HuangPublished in: AAAI (2021)
Keyphrases
- support vector
- distributed systems
- cooperative
- training set
- data structure
- motor skills
- training process
- distributed environment
- mobile agents
- error rate
- test set
- training sessions
- peer to peer
- neural network
- fault tolerant
- echo state networks
- training samples
- lightweight
- error bounds
- training algorithm
- feedforward neural networks
- gradient information
- supervised learning
- reinforcement learning