FedDM: Data and Model Heterogeneity-Aware Federated Learning via Dynamic Weight Sharing.
Leming ShenYuanqing ZhengPublished in: ICDCS (2023)
Keyphrases
- prior knowledge
- learning models
- data sets
- experimental data
- probability distribution
- learning algorithm
- input data
- image data
- data sources
- probabilistic model
- sensory data
- human experts
- data processing
- hidden variables
- bayesian methods
- data mining
- learning scheme
- distributed data
- accurate models
- raw data
- simulation data
- test data
- computational model
- learning phase
- multiple users
- measured data
- temporal dependencies
- high dimensional data
- high level
- training data
- data collection
- supervised learning
- data analysis
- knowledge discovery
- database
- neural network
- objective function
- data structure
- learning process
- end users
- mathematical model
- data mining techniques
- access control
- learning tasks
- statistical methods
- background knowledge