Table 7 Comparison between FedAvg and FedSGD.

From: Federated learning with differential privacy for breast cancer diagnosis enabling secure data sharing and model integrity

Feature

FedAvg

FedSGD

Local updates performed

Yes (multiple per round)

No (only one per round)

Bandwidth usage

Lower

Higher

Convergence stability

More stable

Higher variance in updates

Communication frequency

Less frequent

More frequent

Preferred for FL?

Yes

No

Back to article page