federated learning protocol
/ˌfed.əˈreɪ.tɪd ˈlɜː.nɪŋ ˈprəʊ.tə.kɒl/A federated learning protocol is a set of rules and steps for training a machine learning model across many devices or computers without moving all the data to one place. Each device learns locally, then sends model updates instead of raw data.
- The app uses a federated learning protocol to protect user data.
- Engineers tested the federated learning protocol on hospital phones.
- This protocol lets devices learn together without sharing private files.
Adinary Nuance
A federated learning protocol is more specific than federated learning. Federated learning is the whole training method, while the protocol is the rule set that makes it work. It is also narrower than framework, which can mean a bigger system with tools and code. Use protocol when you mean the agreed process or communication rules.
In other languages
- Vietnamese
- giao thức học liên kết
- Spanish
- protocolo de aprendizaje federado
- Chinese
- 联邦学习协议
- Japanese
- 連合学習プロトコル
- Korean
- 연합 학습 프로토콜
Etymology
This term comes from federated, from Latin foederare, meaning “to unite by agreement,” and learning protocol, a modern computer-science phrase. It became common in the 2010s with privacy-focused machine learning.
Common phrases
Synonyms
Related words
Frequently asked questions
- Is federated learning protocol a common phrase?
- It is common in academic and technical writing, but not in everyday speech.
- What is the difference between federated learning and federated learning protocol?
- Federated learning is the method. The protocol is the set of rules used in that method.
- Is this phrase used in business writing?
- Yes, especially in tech, AI, and privacy-related documents.