Research question
Can models be trained across devices without centralising data?
Key findings
- Federated averaging trains shared models from local updates with far fewer communication rounds
Home/Research Paper Library/Communication-Efficient Learning of Deep Networks from Decentralized Data
Communication-Efficient Learning of Deep Networks from Decentralized Data
Introduced federated learning, the framework for training on decentralised data whose own attack surface (gradient leakage, backdoors) became a research field.
Contribute
Spot an error, or know a record we should add? Members can propose additions and submit corrections.