Edge 281: Cross-Device Federated Learning
Cross device federated learning(FL), Google's work on FL with differential privacy and the FedLab framework
In this issue:
The concept of cross-device federated learning.
Google’s research about federated learning with differential privacy guarantees.
The FedLab framework.
💡 ML Concept of the Day: Cross-Device Federated Learning
Continuing our series about federated learning, we would like to cover one of the most typical architectures. Cross-device federated learning(CDFL) could be considered a generalization of Google’s initial federated learning architecture based on a large federation of individual devices. This approach contrasts with cross-silo federated learning(CSFL), explored last week, in which organizations join the federation through established silos or nodes. A basic comparison between CDFL and CSFL reveals some key differences.