Back to issue
Share on :
Viewed 91 times
28 June 2024

FEDERATED LEARNING, a promising advance for healthcare data processing

Published by Maureen Clerc , Aurélien Bellet and Marco Lorenzi | N° 213 - AI & Healthcare & Digital & Defense

Federated learning, an innovative approach to machine learning that makes it possible to process sensitive data without sharing it directly, offers very interesting prospects in the healthcare field. It could enable major advances in medical research and public health, by reconciling the exploitation of data by artificial intelligence with the protection of patient privacy. However, this method also presents technical, regulatory and organizational challenges.


Learning from data without having access to it Federated learning is a decentralized approach to artificial intelligence where a model is trained on data held by different entities (hospitals, companies, connected objects), without this data leaving its original environment. This method is particularly interesting for the analysis of sensitive data such as medical data, which cannot be shared due to legal and ethical constraints. Since its emergence around 2016-2017, federated learning has been...
This article is reserved for subscribers to the magazine

Authors

Maureen Clerc
Aurélien Bellet
Marco Lorenzi
  Previous article Next article  

Articles in this issue


Latest issues published
Most popular articles
Most popular tags