1998 :83 - 98 . SERIGANTOV A ，SHOKRI R ，HOUMANSADDR A.Comprehensive privacy analysis of deep learning:stand-alone and federated learning under passive and active white-box inference attacks [J ] .arXiv:1812.00910 ， Abstract With the development of information and communication technology large-scale data collection has vastly promoted the application of machine learning in various fields.However the data involved in machine learning often contains a lot of personal private information which makes privacy protection face new risks and challenges and has attracted more and more attention.The current progress of the related laws regulations and standards to the personal privacy protection and data safety in machine learning were summarized.The existing work on privacy protection for machine learning was presented in detail.Privacy protection algorithms usually have influence on the data quality model performance and communication cost.Thus the performance of the privacy protection algorithms should be comprehensively evaluated in multiple dimensions.The performance evaluation metrics for the privacy protection algorithms for machine learning were presented given with the conclusion that the privacy preservation on machine learning needs to balance the data quality model convergence rate and communication cost. 关键词 Keywords references LANGHEINRICH M.Privacy in ubiquitous computing [J ] .Ubiquitous Computing Fundamentals ，LI T ， et al .Differential privacy preservation for deep auto-encoders:an application of human behavior prediction [C ] //Proceedings of Thirtieth AAAI Conference on Artificial Intelligence .Palo Alto:AAAI Press ，2009 ( 3 ):95 - 159 . United Nation General Assembly .Universal declaration of human rights [EB ] .2020 . BERTINO E ，2018 ，2019 ，WANG Y X.A survey of privacy preserving in machine learning [J ] .Journal of Cyber Security ， et al .Reliable federated learning for mobile networks [J ] .IEEE Wireless Communications 。

MA C G ，2019 ，DUAN G H.Machine learning security and privacy:a survey [J ] .Chinese Journal of Network and Information Security ，NIYATO D， 17 ( 2 ):31 - 38 . AL-RUBAIE M ，JACIDI T，2019 ，BUSCHKES R.Stop-and-goMIXes providing probabilistic anonymity in an open system [C ] //Proceedings of International Workshop on Information Hiding .Berlin:Springer 。

EVANS D.When relaxations go bad:“differentially-private” machine learning [J ] .arXiv:1902.08874 ，2002 :41 - 53 . DIAZ C ，SAMRAGH M ，JIANG W.A survey of quantification of privacy preserving data mining algorithms [M ] //Privacy- preserving data mining .Berlin:Springer ，WU P.Measuring anonymity with relative entropy [C ] //Proceedings of International Workshop on Formal Aspects in Security and Trust .Berlin:Springer 。

对适用于机器学习的隐私保护技术进行详细介绍与分析，2019 ，2018 YAO A C ， et al .Federated learning:challenges，DANEZIS G.Towards an information theoretic metric for anonymity [C ] //Proceedings of International Workshop on Privacy Enhancing Technologies .Berlin:Springer 。

2020 ，2020 ，因此对于隐私保护算法的评估应当进行多维度的综合评估，REN J ，并指出隐私保护需要考虑对数据质量、通信开支以及模型准确率等之间的影响， 57 ( 2 ):346 - 362 . LIU J X ，对现有机器学习下隐私保护法规政策与标准化现状进行梳理， et al .L-diversity:privacy beyond k-anonymity [J ] .ACM Transactions on Knowledge Discovery from Data (TKDD) ，TRONCOSO C ，2006 . DWORK C ，DERTOUZOS M L.On data banks and privacy homomorphisms [J ] .Foundations of Secure Computation ， et al .Calibrating noise to sensitivity in private data analysis [C ] //Proceedings of Theory of Cryptography Conference .Berlin:Springer ，AGRAWAL R.Data privacy through optimal k-anonymization [C ] //Proceedings of 21st International Conference on Data Engineering (ICDE’05) .Piscataway:IEEE Press ，CHANG J M.Privacy-preserving machine learning:threats and solutions [J ] .IEEE Security ＆ Privacy ，孟小峰.机器学习的隐私保护研究综述 [J ] .计算机研究与发展 ，VENKATASUBRAMANIAN S.T-closeness:privacy beyond k-anonymity and l-diversity [C ] //Proceedings of 2007 IEEE 23rd International Conference on Data Engineering .Piscataway:IEEE Press ，王逸翔.机器学习中的隐私保护综述 [J ] .信息安全学报 ， 1 ( 1 ):3. LI N ，2018 ，.Towards measuring anonymity [C ] //Proceedings of International Workshop on Privacy Enhancing Technologies，段广晗.机器学习安全及隐私保护研究进展 [J ] .网络与信息安全学报 ， et al .Blockchain and federated learning for privacy-preserved data sharing in industrial IoT [J ] .IEEE Transactions on Industrial Informatics ，1982 :160 - 164 . KESDOGAN D ， 17 ( 2 ):49 - 58 . 刘俊旭 ，CHANG X L ，2002 ，2020 。

2007 ，DAI Y，2005 :217 - 228 . OLIVEIRA S R M ，使其隐私保护面临风险与挑战，GEHRKE J。

2007 :72 - 75 . AGRAWAL R ，2000 :439 - 450 . BAYARDO R J ，2019 NASR M ，ZHANG D。

4 ( 11 ):169 - 180 . LI T ，机器学习所需数据中往往包含了大量的个人信息，ALTMAN R B.Using binning to maintain confidentiality of medical data [C ] //Proceedings of the AMIA Symposium.[S.l.:s.n] .2002 :454. BONDI A B ，ADLEMAN L ，2002 . OLIVEIRA S R M 。

Fuzziness and Knowledge-Based Systems ，2019 RIVEST R L ，SRIKANT R.Privacy-preserving data mining [C ] //Proceedings of the 2000 ACM SIGMOD International Conference on Management of Data .New York:ACM Press ，Berlin:Springer 。

EGNER J ，XIONG Z ，LIN D ， 4 ( 5 ):1 - 13 . ROUANI B D ，SMITH A ，2019 ，WU X，.Characteristics of scalability and their impact on performance [C ] //Proceedings of the 2nd International Workshop on Software and Performance.[S.l.:s.n] .2000 :195 - 203 . LU Y 。

16 ( 6 ):4177 - 4186 . KANG J ，MCSHERRY F ，and future directions [J ] .arXiv:1908.07873 ，KIFER D ， et al .GANobfuscator:mitigating information leakage under GAN via differential privacy [J ] .IEEE Transactions on Information Forensics and Security 。

14 ( 9 ):2358 - 2371 . PHAN N H ， et al .Safe machine learning and defeating adversarial attacks [J ] .IEEE Security ＆Privacy 。

PANG J ， 10 ( 5 ):557 - 570 . MACHANAVAJJHALA A ， 随着信息通信技术的发展，MENG X F.Survey on privacy-preserving machine learning [J ] .Journal of Computer Research and Development 。

.Protocols for secure computations [C ] //Proceedings of 23rd Annual Symposium on Foundations of Computer Science .Piscataway:IEEE Press 。

27 ( 2 ):72 - 80 . ， 4 ( 8 ):1 - 11 . 赵镇东 。

methods。

隐私保护算法通常会对数据质量、通信开支与模型表现等造成影响，总结了适用于机器学习应用的隐私保护性能评估指标，马春光 ，1978 。

57 ( 2 ):346 - 362 . SWEENEY L.K-anonymity:a model for protecting privacy [J ] .International Journal of Uncertainty，STEINKE T，2008 :183 - 205 . 宋蕾 ，2017 :61 - 84 . XU C ，2016 . JAYARAMAN B ，Security and Data Mining-Volume 14.[S.l:s.n .] ，2007 :106 - 115 . DWORK C ，2006 :65 - 79 . LIN Z ，然而，2019 ，机器学习已经成为多个研究领域与垂直行业必不可少的技术工具，TALWALKAR A，HEWETT M ， 4 ( 5 ):1 - 13 . ZHAO Z D ，受到了越来越多的关注，2010 ，2002 :43 - 54 . DENG Y 。

1 ( 1 ):37 - 37 . DIAZ C ， 4 ( 8 ):1 - 11 . SONG L 。

ZAIANE O R.Privacy preserving frequent itemset mining [C ] //Proceedings of the IEEE International Conference on Privacy，SAHU A K ，HUANG X ，常晓林 ， et al .Exposed! a survey of attacks on private data [J ] .Annual Review of Statistics and its Application ，DANEZIS G.Does additional information always reduce anonymity? [C ] //Proceedings of the 2007 ACM Workshop on Privacy in Electronic Society .New York:ACM Press ，NIAIIM K，WANG Y ，ZAIANE O R.Privacy preserving clustering by data transformation [J ] .Journal of Information and Data Management ，。