Table 1 Comparison between federated system and centralized system.

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

Aspect

Federated system

Centralized system

Data Storage

Local devices (distributed)

Centralized database

Data Exchange

No raw data exchange, only model updates

Raw data is continuously sent to the central server

Privacy

High (data remains on device)

Low (data stored centrally, risk of breaches)

Communication

Optimized for model updates

High bandwidth needed for data transfer

Performance

Slightly lower due to decentralized data

High, as the model accesses the complete dataset

Use Cases

Healthcare, finance, IoT devices

E-commerce, social media, large-scale data analytics

Back to article page