What is Federated Learning?

Federated Learning (FL), also known as collaborative learning, is a decentralized approach to training machine learning models. Unlike centralized methods, which collect data from individual devices to a central server, federated learning keeps data on each device, ranging from phones and laptops to wearables, vehicles, and IoT devices. The model is trained locally on each device using the data stored there, and only the model updates (or learned patterns) are sent to a central server. These updates are aggregated to improve the global model without transferring raw data.

Federated learning is particularly valuable in privacy-sensitive applications, such as personalized recommendations or predictive text. A real-life application is Google’s use of federated learning to enhance next-word prediction models in its Gboard app for Android.