面向群智感知隐私保护的联邦典型相关分析方法

doi:10.12341/jssms22106

系统科学与数学 ›› 2022, Vol. 42 ›› Issue (11): 2859-2873.DOI: 10.12341/jssms22106

• •

面向群智感知隐私保护的联邦典型相关分析方法

李文平, 朱荷蕾

嘉兴学院信息科学与工程学院, 嘉兴 314001

收稿日期:2022-02-18 修回日期:2022-07-12 发布日期:2022-12-13
通讯作者: 李文平, Email: liwenping@hrbeu.edu.cn
基金资助:
教育部产学合作协同育人项目(202101014039),嘉兴市科技特派员专项项目(K2022A015)资助课题.}%如 *国家自然科学基金(10171074)资助课题.

PDF ( 30 ) 引用

李文平, 朱荷蕾. 面向群智感知隐私保护的联邦典型相关分析方法[J]. 系统科学与数学, 2022, 42(11): 2859-2873.

LI Wenping, ZHU Helei. A Federal-Based Canonical Correlation Analysis Approach for Privacy Preserving in Crowdsensing[J]. Journal of Systems Science and Mathematical Sciences, 2022, 42(11): 2859-2873.

A Federal-Based Canonical Correlation Analysis Approach for Privacy Preserving in Crowdsensing

LI Wenping, ZHU Helei

College of Information Science and Engineering, Jiaxing University, Jiaxing

Received:2022-02-18 Revised:2022-07-12 Published:2022-12-13

群智感知数据往往携带了丰富的敏感信息, 联邦学习技术的兴起为解决感知系统之间隐私安全的协作计算提供了一条可行途径.针对群智感知关联性探测过程中可能导致的隐私泄露问题, 提出一种联邦场景下支持隐私保护的典型相关分析方法.该方法通过构造联邦特征并基于非线性随机耦合策略进行特征保护,将典型相关分析求解过程分解为两个完全独立的运算分别由联邦端和服务器端执行,在数据级、运算级和特征级上对感知数据进行多层次保护.实验结果表明,文章方法对联邦场景的适用性和感知数据的保护性都较好.

关键词： 群智感知, 隐私保护, 联邦学习, 典型相关分析

The crowdsensing data often carries a wealth of sensitive information. The rise of federated learning technology provides a feasible way to solve the privacy preserving problem of cooperative computing between crowdsensing systems. A canonical correlation analysis approach, preserving privacy and supporting federated scenarios, is proposed to solve the problem of privacy disclosure in the process of correlation detection on crowdsensing. By constructing federal eigens and preserving them based on nonlinear random coupling strategy, the proposed approach decomposes the solving process of canonical correlation analysis into two completely independent operations, which are executed on the federals and the server respectively. The privacy of crowdsensing is preserved at the data-level, operation-level and eigen-level effectively. Experimental results verify the effectiveness of the proposed approach in both of applicability to federated scenarios and ability to preserve crowdsensing.

Key words: Crowdsensing, privacy preserving, federated learning, canonical correlation analysis

MR(2010)主题分类:

62G35

分享此文:

[1] Guo B, Wang Z, Yu Z W, et al. Mobile crowd sensing and computing: The review of an emerging human-powered sensing paradigm. ACM Computing Surveys, 2015, 48(1): 1-31.
[2] 孙佳琛, 王金龙, 陈瑾, 等. 群体智能协同通信:愿景、模型和关键技术. 中国科学:信息科学, 2020, 50(3): 307-317. (Sun J C, Wang J L, Chen J, et al. Cooperative communication based on swarm intelligence: Vision, model, and key technology. Scientia Sinica Informationis, 2020, 50(3): 307-317.)
[3] 安健, 彭振龙, 桂小林, 等. 群智感知中基于公交系统的任务分发机制研究. 计算机学报, 2019, 42(2): 65-78. (An J, Peng Z L, Gui X L, et al. Research on task distribution mechanism based on public transit system in crowd sensing. Chinese Journal of Computers, 2019, 42(2): 65-78.)
[4] 景瑶, 郭斌, 陈荟慧, 等. CrowdTracker: 一种基于移动群智感知的目标跟踪方法. 计算机研究与发展, 2019, 56(2): 328-337. (Jing Y, Guo B, Chen H H, et al. CrowdTracker: Object tracking using mobile crowd sensing. Journal of Computer Research and Development, 2019, 56(2): 328-337.)
[5] 王健, 刘嘉欣, 赵国生, 等. 移动群智感知中基于协同排序的任务推荐方法. 电子学报, 2021, 49(10): 2012-2019. (Wang J, Liu J X, Zhao G S, et al. Task recommendation method based on collaborative ranking in mobile crowd sensing. Acta Electronica Sinica, 2021, 49(10): 2012-2019.)
[6] Warnat H S, Schultze H, Shastry K L, et al. Swarm learning for decentralized and confidential clinical machine learning. Nature, 2021, 594(7862): 265-270.
[7] 熊金波, 毕仁万, 田有亮, 等. 移动群智感知安全与隐私:模型、进展与趋势. 计算机学报, 2021, 44(9): 1949-1966. (Xiong J B, Bi R W, Tian Y L, et al. Security and privacy in mobile crowdsensing: Models, progresses, and trends. Chinese Journal of Computers, 2021, 44(9): 1949-1966.)
[8] Vahdat N H, Asani E, Mahmoodian Z, et al. Context-aware computing for mobile crowd sensing: A survey. Future Generation Computer Systems - The International Journal of Escience, 2019, 99(12): 321-332.
[9] 史鼎元, 王晏晟, 郑鹏飞, 等. 面向企业数据孤岛的联邦排序学习. 软件学报, 2021, 32(3): 669-688. (Shi D Y, Wang Y S, Zheng P F, et al. Cross-silo federated learning-to- rank. Journal of Software, 2021, 32(3): 669-688.)
[10] 周传鑫, 孙奕, 汪德刚, 等. 联邦学习研究综述. 网络与信息安全学报, 2021, 7(5): 1-16. (Zhou C X, Sun Y, Wang D G, et al. Survey of federated learning research. Chinese Journal of Network and Information Security, 2021, 7(5): 1-16.)
[11] Zhou B, Chen L, Zhao S, et al. Spatio-temporal analysis of urban crime leveraging multisource crowdsensed data. Personal and Ubiquitous Computing, 2021, 25(1): 1-14.
[12] Marchang N, Tripathi R. KNN-ST: Exploiting spatio-temporal correlation for missing data inference in environmental crowd sensing. IEEE Sensors Journal, 2021, 21(3): 3429-3436.
[13] Hotelling H. Relations between two sets of variates. Biometrika, 1936, 28(3/4): 321-377.
[14] 陈志文, 李学明, 徐绍龙, 等. 基于特征相关性的牵引系统主回路接地故障诊断. 自动化学报, 2021, 47(7): 1516-1529. (Chen Z W, Li X M, Xu S L, et al. Feature correlation-based ground fault diagnosis method for main circuit of traction system. Acta Automatica Sinica, 2021, 47(7): 1516-1529.)
[15] Huang P, Lu J. Learning trajectory patterns via canonical correlation analysis. International Journal of Cognitive Informatics and Natural Intelligence, 2021, 15(2): 1-17.
[16] 周航星, 陈松灿. 有序判别典型相关分析. 软件学报, 2014, 25(9): 2018-2025. (Zhou H X, Chen S C. Ordinal discriminative canonical correlation analysis. Journal of Software, 2014, 25(9): 2018-2025.)
[17] Pehlevan C, Zhao X, Sengupta A M, et al. Neurons as canonical correlation analyzers. Frontiers in Computational Neuroscience, 2020, 14(1): 55.
[18] 王宇霞, 赵清杰, 赵留军. 基于FREAK和P3CA的鲁棒目标跟踪. 计算机学报, 2015, 38(6): 1188-1201. (Wang Y X, Zhao Q J, Zhao L J. Robust object tracking based on FREAK and P3CA. Chinese Journal of Computers, 2015, 38(6): 1188-1201.)
[19] 刘建伟, 孙正康, 刘泽宇, 等. 核典型关联性分析相关特征提取与核逻辑斯蒂回归域自适应学习. 电子学报, 2016, 44(12): 2908-2915. (Liu J W, Sun Z K, Liu Z Y, et al. Domain adaptation learning with kernel logistic regression and kernel canonical correiation anaiysis. Acta Electronica Sinica, 2016, 44(12): 2908-2915.)
[20] 孙平, 徐宗本, 申建中. 基于核化原理的非线性典型相关判别分析. 计算机学报, 2004, 27(6): 789-795. (Sun P, Xu Z B, Shen J Z. Nonlinear canonical correlation analysis for discrimination based on kernel methods. Chinese Journal of Computers, 2004, 27(6): 789-795.)
[21] 杨静, 李文平, 张健沛. 基于秩2更新的多维数据流典型相关跟踪算法. 电子学报, 2012, 40(9): 1765-1774. (Yang J, Li W P, Zhang J P. A tracking algorithm based on rank two modifications for canonical correlation analysis of multidimensional data streams. Acta Electronica Sinica, 2012, 40(9): 1765-1774.)
[22] 侯书东, 孙权森. 稀疏保持典型相关分析及在特征融合中的应用. 自动化学报, 2012, 38(4): 659-665. (Hou S D, Sun Q S. Sparsity preserving canonical correlation analysis with application in feature fusion. Acta Automatica Sinica, 2012, 38(4): 659-665.)
[23] 潘玉, 陈晓红, 李舜酩, 等. 块增量典型相关分析. 计算机科学与探索, 2022, 16(8): 1809-1818. (Pan Y, Chan X H, Li S M, et al. Chunk incremental canonical correlation analysis. Journal of Frontiers of Computer Science and Technology, 2022, 16(8): 1809-1818.)
[24] 姜云卢, 邓罡, 文诗涵, 等. 高维稳健典型相关分析研究与应用. 系统科学与数学, 2021, 41(10): 2965-2976. (Jiang Y L, Deng G, Wen S H, et al. Research and application of high-dimension robust canonical correlation analysis. Journal of Systems Science and Mathematical Sciences, 2021, 41(10): 2965-2976.)
[25] Wang W, Zhou Y H. Eigenvector-based sparse canonical correlation analysis: Fast computation for estimation of multiple canonical vectors. Journal of Multivariate Analysis, 2021, 185: 104781.
[26] Chen J, Wang G, Giannakis G B. Graph multiview canonical correlation analysis. IEEE Transactions on Signal Processing, 2019, 67(11): 2826-2838.
[27] Kaloga Y, Borgnat P, Chepuri S P, et al. Variational graph autoencoders for multiview canonical correlation analysis. Signal Processing, 2021, 188: 108182.
[28] 苏树智, 谢军, 平昕瑞, 等. 图强化典型相关分析及在图像识别中的应用. 电子与信息学报, 2021, 43(11): 3342-3349. (Su S Z, Xie J, Ping X R, et al. Graph enhanced canonical correlation analysis and its application to image recognition. Journal of Electronics and Information Technology, 2021, 43(11): 3342-3349.)
[29] 许欢, 苏树智, 颜文婧, 等. 面向图像识别的测地局部典型相关分析方法. 电子与信息学报, 2020, 42(11): 2813-2818. (Xu H, Sun S Z, Yan W J, et al. A geodesic locality canonical correlation analysis method for image recognition. Journal of Electronics and Information Technology, 2020, 42(11): 2813-2818.)
[30] Gloaguen A, Philippe C, Frouin V, et al. Multiway generalized canonical correlation analysis. Biostatistics, 2022, 23(1): 240-256.
[31] Hardoon D R, Szedmak S, Shawe T J. Canonical correlation analysis: An overview with application to learning methods. Neural Computation, 2004, 16(12): 2639-2664.
[32] Ewerbring L M, Luk F T. Canonical correlations and generalized SVD: Applications and new algorithms. Journal of Computational and Applied Mathematics, 1989, 27(1-2): 37-52.
[33] Golub G H. Matrix decompositions and statistical calculations. Statistical Computation. New York: Academic Press, 1969, 365-397.

[1]	姜云卢, 邓罡, 文诗涵, 刘峻成. 高维稳健典型相关分析研究与应用[J]. 系统科学与数学, 2021, 41(10): 2965-2976.
[2]	洪志敏，闫在在. 估计敏感特征比例和敏感指标均值的问卷调查技术[J]. 系统科学与数学, 2012, 32(3): 344-354.

阅读次数

全文

摘要