Federated Learning
Definition
Federated learning is a machine learning technique that trains an algorithm across multiple decentralized edge devices or servers holding local data samples, without exchanging their data samples. This approach stands in contrast to traditional centralized machine learning techniques where all the local datasets are uploaded to one server.
Why It Matters
Federated learning allows for AI models to be trained on large amounts of data without compromising user privacy. It is a powerful technique for building smart, personalized features without needing to collect sensitive data on a central server.
Contextual Example
Your phone's keyboard might use federated learning to improve its next-word prediction. It trains a small model on your local typing data. It then sends an anonymized summary of the model's updates to a central server, which aggregates these updates from thousands of users to improve the global prediction model, without ever seeing your actual messages.
Common Misunderstandings
- It is a "train local, aggregate global" model.
- It is a key technology for privacy-preserving machine learning.