Can hierarchical client clustering mitigate the data heterogeneity effect in federated learning?
Seungjun LeeMiri YuDaegun YoonSangyoon OhPublished in: IPDPS Workshops (2023)
Keyphrases
- data sets
- data sources
- synthetic data
- data analysis
- learning algorithm
- data collection
- data processing
- data mining algorithms
- learning tasks
- prior knowledge
- raw data
- learning process
- data mining tasks
- training data
- original data
- learning models
- learned models
- spatial data
- anomaly detection
- central server
- neural network
- online learning
- knowledge discovery
- data points
- input data
- learning systems
- missing data
- end users
- active learning
- xml documents
- k means
- spectral clustering
- digital libraries
- data streams
- data structure
- machine learning