FedGSync: Jointly Optimized Weak Synchronization and Gradient Transmission for Fast Distributed Machine Learning in Heterogeneous WAN.
Huaman ZhouYihong HeZhihao ZhangLong LuoHongfang YuGang SunPublished in: SMC (2023)
Keyphrases
- machine learning
- distributed information systems
- distributed environment
- distributed systems
- cooperative
- heterogeneous environments
- heterogeneous data
- distributed data
- computer vision
- distributed heterogeneous
- reinforcement learning
- loosely coupled
- wide area network
- heterogeneous databases
- data mining
- data transmission
- transparent access
- machine learning algorithms
- text mining
- natural language processing
- information extraction
- multi agent
- machine learning methods
- decision trees
- distributed data sources
- artificial intelligence
- remote sites
- heterogeneous systems
- active learning
- distributed processing
- heterogeneous data sources
- explanation based learning
- distributed search
- peer to peer
- learning tasks
- feature selection