A machine learning approach where models are trained across multiple devices without sharing raw data. Used in privacy-sensitive AI applications like mobile and healthcare.
Example:
Google uses federated learning to improve predictive text on Android devices without uploading user data.
