已入深夜,您辛苦了!由于当前在线用户较少,发布求助请尽量完整地填写文献信息,科研通机器人24小时在线,伴您度过漫漫科研夜!祝你早点完成任务,早点休息,好梦!

Differentially private cross-silo federated learning

作者
Mikko Heikkilä,Antti Koskela,Kana Shimizu,Samuel Kaski,Antti Honkela
出处
期刊:Cornell University - arXiv [Cornell University]
被引量:22
标识
DOI:10.48550/arxiv.2007.05553
摘要

Strict privacy is of paramount importance in distributed machine learning. Federated learning, with the main idea of communicating only what is needed for learning, has been recently introduced as a general approach for distributed learning to enhance learning and improve security. However, federated learning by itself does not guarantee any privacy for data subjects. To quantify and control how much privacy is compromised in the worst-case, we can use differential privacy. In this paper we combine additively homomorphic secure summation protocols with differential privacy in the so-called cross-silo federated learning setting. The goal is to learn complex models like neural networks while guaranteeing strict privacy for the individual data subjects. We demonstrate that our proposed solutions give prediction accuracy that is comparable to the non-distributed setting, and are fast enough to enable learning models with millions of parameters in a reasonable time. To enable learning under strict privacy guarantees that need privacy amplification by subsampling, we present a general algorithm for oblivious distributed subsampling. However, we also argue that when malicious parties are present, a simple approach using distributed Poisson subsampling gives better privacy. Finally, we show that by leveraging random projections we can further scale-up our approach to larger models while suffering only a modest performance loss.

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
Taibeile发布了新的文献求助10
刚刚
深情安青应助tt采纳,获得10
1秒前
1秒前
4秒前
大国发布了新的文献求助10
4秒前
霄泽发布了新的文献求助10
5秒前
molihuakai应助Nl采纳,获得10
6秒前
112233完成签到 ,获得积分10
8秒前
11秒前
12秒前
鹏宝完成签到,获得积分10
14秒前
端庄白易完成签到,获得积分10
14秒前
上官若男应助跳跃白竹采纳,获得10
15秒前
tt发布了新的文献求助10
16秒前
海荷完成签到,获得积分10
16秒前
笑点低的丹蝶完成签到,获得积分10
16秒前
16秒前
严兴明完成签到,获得积分10
16秒前
852应助大国采纳,获得10
16秒前
Fan完成签到 ,获得积分0
16秒前
bluebell完成签到,获得积分10
17秒前
18秒前
鹏宝发布了新的文献求助10
18秒前
Gy发布了新的文献求助10
21秒前
Nl完成签到,获得积分10
28秒前
斯文的野狼完成签到,获得积分10
29秒前
30秒前
哈哈发布了新的文献求助10
31秒前
33秒前
五邑大学完成签到 ,获得积分10
34秒前
秦不语发布了新的文献求助10
40秒前
43秒前
CipherSage应助宁燕采纳,获得10
44秒前
lii完成签到,获得积分10
46秒前
47秒前
47秒前
50秒前
52秒前
54秒前
宁燕发布了新的文献求助10
55秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
Weaponeering: An Introduction Fourth Edition, Volume 1 1000
Advanced Weaponeering Fourth Edition, Volume 2 1000
Evidence Summary. Injection (subcutaneous):op- timal administration 1000
悉尼大学博士学位论文,题目:Modelling and testing of one-sided stitched laminated composites. 作者:Kristopher P. Plain 700
Matrix Methods in Data Mining and Pattern Recognition Second Edition 610
Curating Socialism: A Handbook of International Art Exhibitions 1947-1989 550
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 工程类 有机化学 化学工程 生物化学 计算机科学 内科学 物理 复合材料 催化作用 细胞生物学 无机化学 光电子学 物理化学 电极 基因
热门帖子
关注 科研通微信公众号,转发送积分 7496135
求助须知:如何正确求助?哪些是违规求助? 9087104
关于积分的说明 19381985
捐赠科研通 7107349
什么是DOI,文献DOI怎么找? 3249980
关于科研通互助平台的介绍 2419376
邀请新用户注册赠送积分活动 2235736