Meaning of Federated Learning

Simple definition

Federated learning is a machine learning technique where multiple devices train a shared model collaboratively without sharing their data, preserving privacy and reducing data transfer.

How to use Federated Learning in a professional context

It’s particularly useful in privacy-sensitive applications like healthcare, finance, and mobile devices, where data cannot leave local environments.

Concrete example of Federated Learning

Smartphones participate in federated learning to improve a language model for predictive text without sharing users’ typing data.