【论文分享】DASFFA 2019 | FedSel: 基于Top-k维度选择的本地差分隐私联邦学习算法

发布人

Paper:《FedSel: Federated SGD under Local Differential Privacy with Top-k Dimension Selection》
Abstract: As massive data are produced from small gadgets, federated learning on mobile devices has become an emerging trend. In the federated setting, Stochastic Gradient Descent (SGD) has been widely used in federated learning for various machine learning models. To prevent privacy leakages from gradients that are calculated on users&#39; sensitive data, local differential privacy (LDP) has been considered as a privacy guarantee in federated SGD recently. However, the existing solutions have a dimension dependency problem: the injected noise is substantially proportional to the dimension d. In this work, we propose a two-stage framework FedSel for federated SGD under LDP to relieve this problem. Our key idea is that not all dimensions are equally important so that we privately select Top-k dimensions according to their contributions in each iteration of federated SGD. Specifically, we propose three private dimension selection mechanisms and adapt the gradient accumulation technique to stabilize the learning process with noisy updates. We also theoretically analyze privacy, accuracy and time complexity of FedSel, which outperforms the state-of-the-art solutions. Experiments on real-world and synthetic datasets verify the effectiveness and efficiency of our framework.

打开封面下载高清视频观看高清视频视频下载器

【论文分享】DASFFA 2019 | FedSel: 基于Top-k维度选择的本地差分隐私联邦学习算法

【论文分享】差分隐私~《deep learning with differential privacy》~Moments Accoutant的关键思想

【论文分享】特邀西安电子科技大学博士分享《Local and Central DP for Robustness and Privacy in FL》

【论文分享】《基于本地差分隐私的多维数据收集和分析》 | ICDE 2019

【论文分享】湖南大学江政良分享论文《通过LWE难题实现高效且满足差分隐私的联邦学习安全聚合》| usenix security'22

【论文分享】中科院博士刘博超论文分享《DPGEN：用于自然图像合成的差分隐私生成能量引导网络》

【论文分享】《资源受限边缘计算系统中的自适应联邦学习》| JSAC（CCF A）

【论文分享】VLDB 2024 | 基于选择性更新和释放的差分隐私梯度下降算法

【ASTAPLE DP Group】SIGMO 2022 | LDP-IDS：无限数据流的本地差分隐私保护 - Rong Du

【论文讨论】《Learning Differenitally Private Languange Model》~Client Level FL-DP

【论文分享】《Locally Differentially Private Protocols for Frequency Estimation》

【论文分享】基于K-means 聚类的差分隐私（指数机制）协同过滤推荐算法

【教材分享交流】差分隐私—《Differential Privacy From Theory to Practice》-chapter1、chapter2

【组会汇报】《Hiding Among the Clones: A Simple and Nearly Optimal Analysis of Privacy》

【论文汇报】特邀浙江大学博士冯浩哲（知乎大V“捡到一束光”）分享《KD3A: 一种满足隐私保护要求的去中心化无监督域适应范式》[ICML2021]

【论文分享】《用于个性化联邦学习的分层模型聚合》 | CPVR 2022

【论文分享】《一种分布估计场景下的实用性本地差分隐私机制》| USENIX 2019

【组会汇报】差分隐私-《PATESEMI-SUPERVISED KNOWLEDGE TRANSFER FOR DEEP LEARNING》（PATE）

【学习分享】《本地化差分隐私综述》—LDP

【论文分享】《renyi differential privacy》-瑞丽差分隐私

【ASTAPLE DP Group】SIGMO 2020 | 本地差分隐私下的数值分布估计 - Liantong Yu

【组会分享】《The Privacy Blanket of the Shuffle Model》~隐私毯子

【组会分享】自适应差分隐私深度学习-《An Adaptive and Fast Convergent Approach to DP DL 》

【组会分享】《Gaussian Differential Privacy》

【教材分享】拉普拉斯机制？高斯机制？严格差分隐私？松弛差分隐私？

【论文分享】《基于DPSGD的图节点分类任务》| ICLR 2023 （欢迎有做图神经/图联邦结合DP的同僚与我联系交流）

【组会论文记录】《User-Level Privacy-Preserving Federated Learning: Analysis and Perform》

【论文分享】特邀中科院博士蒋雪枫最新论文《基于局部自正则化的抗噪声标签联合学习》CIKM 2022！

【ASTAPLE DP Group】Advanced Composition in Differential Privacy - Xun Ran

【ASTAPLE DP Group】CSF 2017 | Renyi Differential Privacy (瑞丽差分隐私) - Jianping Cai

【录制搬运】浙江大学特聘研究员陈超超 《隐私保护机器学习》

【组会分享】《Comprehensive Privacy Analysis of Deep Learning》集中式与联邦场景下的消极与积极的成员推理攻击

【转载】Synthesizing Relational Data with Differential Privacy | Dr. Xiaokui Xia

【论文讨论】《Distributed Gaussian Differentially Privacy Via Shuffing》

【转载】Three Flavors of Differentially Private Federated Learning | Dr. Yang Cao

【论文分享】USENIX 2023 | 利用社区信息进行差分隐私图数据发布

【转载】Differential Privacy: Potential and Limitations | Prof. Ninghui Li

【教材分享】梯度下降在强凸/凸/非凸下的收敛速率证明

【转载】Privacy is Not an Afterthought | Dr. Xi He

【录制搬运】埃默里大学 Li Xiong 教授《Federated Learning: Privacy, Efficiency, and Robustness》

【学术论文写作】英文学术会议论文写作经验分享-Jeff

【录制搬运】浙江大学特聘研究员陈超超《隐私保护机器学习》