FedHiSyn: A Hierarchical Synchronous Federated Learning Framework for Resource and Data Heterogeneity.
Guanghao LiYue HuMiao ZhangJi LiuQuanjun YinYong PengDejing DouPublished in: ICPP (2022)
Keyphrases
- data sets
- data sources
- synthetic data
- database
- image data
- data analysis
- data processing
- distributed data
- data quality
- human experts
- background knowledge
- max margin learning
- heterogeneous sources
- bayesian methods
- learning algorithm
- unsupervised learning
- data collection
- high quality
- data structure
- reinforcement learning
- knowledge acquisition
- learning process
- databases
- knowledge discovery
- experimental data
- data distribution
- learning tasks
- feature selection
- raw data
- learning scheme
- data streams
- database systems
- data mining techniques
- online learning