Towards Efficient and Stable K-Asynchronous Federated Learning With Unbounded Stale Gradients on Non-IID Data.
Zihao ZhouYanan LiXuebin RenShusen YangPublished in: IEEE Trans. Parallel Distributed Syst. (2022)
Keyphrases
- data sources
- background knowledge
- prior knowledge
- small number
- data sets
- distributed data
- raw data
- online learning
- data analysis
- learning process
- database
- data mining techniques
- original data
- spatial data
- learning systems
- statistical analysis
- active learning
- databases
- reinforcement learning
- high quality
- training data
- sensor data
- learned models
- learning models
- noisy data
- data mining
- learning algorithm
- learning tasks
- data distribution
- experimental data
- social networks
- image data
- synthetic data
- knowledge acquisition
- data processing
- data structure
- data points