Findings from University of Electronic Science and Technology of China Yields Ne w Findings on Machine Learning (Improved Gradient Leakage Attack Against Compres sed Gradients In Federated Learning)

中国电子科技大学的研究成果在机器学习（联邦学习中针对压缩梯度的改进梯度泄漏攻击）方面有了新的发现

Robotics & Machine Learning Daily News2024，Issue(Dec.4) ：165-166.

来源：

NETL
NSTL

Findings from University of Electronic Science and Technology of China Yields Ne w Findings on Machine Learning (Improved Gradient Leakage Attack Against Compres sed Gradients In Federated Learning)

中国电子科技大学的研究成果在机器学习（联邦学习中针对压缩梯度的改进梯度泄漏攻击）方面有了新的发现

扫码查看

摘要

由一名新闻记者-机器人与机器学习日报的工作人员新闻编辑每日新闻-关于机器学习的详细数据已经呈现。根据新闻报道源于中华人民共和国四川，由NewsRx记者报道，研究称，“已分发”机器学习通过收集梯度而不是收集梯度来保护隐私培训数据。近年来的研究表明，梯度泄漏法在分布式机械中是可行的学习，也就是说，训练数据可以从共享的梯度中重建。

Abstract

By a News Reporter-Staff News Editor at Robotics & Machine Learning DailyNews Daily News – Data detailed on Machine Learning have been presented. According to news reportingoriginating from Sichuan, People’s Republic of China, by NewsRx correspondents, research stated, “Distributedmachi ne learning, such as federated learning, protects privacy by collecting gradient s instead oftraining data. Recent studies have shown that gradient leakage atta cks are possible in distributed machinelearning, that is, the training data can be reconstructed from the shared gradients.”

Key words

Sichuan/People’s Republic of China/Asi a/Cyborgs/Emerging Technologies/Machine Learning/University of Electronic Sc ience and Technology of China

引用本文复制引用

出版年

2024

Robotics & Machine Learning Daily News

ISSN：

段落导航