What Federated Learning Meaning, Applications & Example

A technique that trains models across devices without centralized data.

What is Federated Learning?

Federated Learning is a machine learning approach that enables model training across multiple devices or servers without sharing raw data. Instead, each device trains a local model on its own data and shares only the model updates (e.g., weights or gradients) with a central server, preserving data privacy and security .

Types of Federated Learning

  1. Horizontal Federated Learning: Used when datasets across different devices have the same features but represent different samples (e.g., multiple hospitals training on patient data with the same attributes).
  2. Vertical Federated Learning: Used when datasets across devices share some overlapping samples but have different features (e.g., a bank and an e-commerce platform collaborating with shared customers).
  3. Federated Transfer Learning: Applies when datasets have different samples and features, allowing knowledge transfer between devices with limited data overlap.

Applications of Federated Learning

Example of Federated Learning

An example of Federated Learning is in smartphone keyboard prediction, where a predictive text model is trained across many users’ devices to improve typing suggestions without centralizing users’ personal data, maintaining privacy while continuously enhancing the model’s accuracy.

Read the Governor's Letter

Stay ahead with Governor's Letter, the newsletter delivering expert insights, AI updates, and curated knowledge directly to your inbox.

By subscribing to the Governor's Letter, you consent to receive emails from AI Guv.
We respect your privacy - read our Privacy Policy to learn how we protect your information.

A

B

C

D

E

F

G

H

I

J

K

L

M

N

O

P

Q

R

S

T

U

V

W

X

Y

Z