Federated learning: Basics and application to the mobile keyboard (ML Tech Talks)

preview_player
Показать описание
In this session of Machine Learning Tech Talks, Research Scientist Françoise Beaufays will discuss the basics and application of the mobile keyboard for federated learning.

Chapters:
0:00 - Introduction
0:43 - Intro to federated learning
4:20 - Mobile keyboards
8:32 - Making federated learning work in practice
23:09 - Applying federated learning to mobile keyboards
31:05 - Where to learn more

Resource:

purpose: Educate, pr_pr: TensorFlow, series: ML Tech Talks, type: DevByte (deck cleanup 10-20min), GDS: Yes,
Рекомендации по теме
Комментарии
Автор

You are an awesome presenter! The explanations are super clear thanks

florianhonicke
Автор

The explanation was clear and well presented

orjihvy
Автор

Thank you so much. Highly recommended vide for learning FL.

muhammadalikhan
Автор

Awesome! Finally some more movement on this since the talk two years ago. I hope you put some effort into making some actionable tutorials — that’s certainly a differentiator to other cloud providers

dinoscheidt
Автор

What are other applications of federated learning outside mobile keyboard application? How intense is the computational speed needed in the mobile device ?

tomasbarrios
Автор

Excellent presentation and a great use of FL. Great balance of privacy and model improvement. Hope to see other keyboards take it up too.

shubhamsaha
Автор

Using customer's device resource for training model seems wrong in my point of view.
Even though by looking at "Devices check in iff".
It's like mining cryptocurrency in others devices.

NaryVip