SPATL: Salient Parameter Aggregation and Transfer Learning for Heterogeneous Clients in Federated Learning.
Sixing YuPhuong NguyenWaqwoya AbebeAli AnwarAli JannesariPublished in: CoRR (2021)
Keyphrases
- transfer learning
- learning tasks
- knowledge transfer
- transfer knowledge
- active learning
- reinforcement learning
- databases
- transferring knowledge
- learning algorithm
- graph transduction
- supervised learning
- multi task learning
- learning process
- domain adaptation
- natural language
- prior knowledge
- training data
- learning problems
- multi task
- feature extraction
- expectation maximization
- previously learned
- natural language processing
- sample selection bias
- data sets
- cross domain learning