🧠
Federated Learning
Privacy
Training a model on devices without sending all raw data to a server.
Definition
Federated learning trains an AI model across many devices so raw data can stay on the device.
In plain English
Training a model on devices without sending all raw data to a server.
Why this matters
Why it matters: It can reduce data sharing, but updates may still leak information if not protected.
Example
Example: A keyboard model learns from many phones without uploading every typed sentence.