On the Protection of Private Information in Machine Learning Systems: Two Recent Approches

The recent, remarkable growth of machine learning has led to intense interest in the privacy of the data on which machine learning relies, and to new techniques for preserving privacy. However, older ideas about privacy may well remain valid and useful. This note reviews two recent works on privacy in the light of the wisdom of some of the early literature, in particular the principles distilled by Saltzer and Schroeder in the 1970s.

[1]  Willis H. Ware,et al.  Security and privacy: similarities and differences , 1967, AFIPS '67 (Spring).

[2]  Yung-Ying Liu,et al.  Privacy and Security in Computer Systems. , 1974 .

[3]  Jerome H. Saltzer,et al.  The protection of information in computer systems , 1975, Proc. IEEE.

[4]  Benedict G. E. Wiedemann.  Protection? , 1998, Science.

[5]  Thomas G. Dietterich.  Ensemble Methods in Machine Learning , 2000, Multiple Classifier Systems.

[6]  Simson L. Garfinkel,et al.  Design principles and patterns for computer systems that are simultaneously secure and usable , 2005 .

[7]  Moni Naor,et al.  Our Data, Ourselves: Privacy Via Distributed Noise Generation , 2006, EUROCRYPT.

[8]  Cynthia Dwork,et al.  Calibrating Noise to Sensitivity in Private Data Analysis , 2006, TCC.

[9]  Sofya Raskhodnikova,et al.  Smooth sensitivity and sampling in private data analysis , 2007, STOC '07.

[10]  Sofya Raskhodnikova,et al.  What Can We Learn Privately? , 2008, 2008 49th Annual IEEE Symposium on Foundations of Computer Science.

[11]  Jerome H. Saltzer,et al.  Principles of Computer System Design: An Introduction , 2009 .

[12]  Bhiksha Raj,et al.  Multiparty Differential Privacy via Aggregation of Locally Trained Classifiers , 2010, NIPS.

[13]  C. Dwork.  A firm foundation for private data analysis , 2011, Commun. ACM.

[14]  Anand D. Sarwate,et al.  Differentially Private Empirical Risk Minimization , 2009, J. Mach. Learn. Res..

[15]  Daniel Kifer,et al.  Private Convex Optimization for Empirical Risk Minimization with Applications to High-dimensional Regression , 2012, COLT.

[16]  Ilya Mironov,et al.  On significance of the least significant bits for differential privacy , 2012, CCS.

[17]  Richard E. Smith,et al.  A Contemporary Look at Saltzer and Schroeder's 1975 Design Principles , 2012, IEEE Security & Privacy.

[18]  Anand D. Sarwate,et al.  Stochastic gradient descent with differentially private updates , 2013, 2013 IEEE Global Conference on Signal and Information Processing.

[19]  Sharon Goldberg,et al.  Calibrating Data to Sensitivity in Private Data Analysis , 2012, Proc. VLDB Endow..

[20]  Yoshua Bengio,et al.  Generative Adversarial Nets , 2014, NIPS.

[21]  Úlfar Erlingsson,et al.  RAPPOR: Randomized Aggregatable Privacy-Preserving Ordinal Response , 2014, CCS.

[22]  Raef Bassily,et al.  Differentially Private Empirical Risk Minimization: Efficient Algorithms and Tight Error Bounds , 2014, 1405.7085.

[23]  Vitaly Shmatikov,et al.  Privacy-preserving deep learning , 2015, 2015 53rd Annual Allerton Conference on Communication, Control, and Computing (Allerton).

[24]  Sergey Ioffe,et al.  Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[25]  Quoc V. Le,et al.  Adding Gradient Noise Improves Learning for Very Deep Networks , 2015, ArXiv.

[26]  Somesh Jha,et al.  Model Inversion Attacks that Exploit Confidence Information and Basic Countermeasures , 2015, CCS.

[27]  Mikhail Belkin,et al.  Learning privately from multiparty data , 2016, ICML.

[28]  Michael Naehrig,et al.  CryptoNets: applying neural networks to encrypted data with high throughput and accuracy , 2016, ICML 2016.

[29]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[30]  Wojciech Zaremba,et al.  Improved Techniques for Training GANs , 2016, NIPS.

[31]  Jeffrey F. Naughton,et al.  Differentially Private Stochastic Gradient Descent for in-RDBMS Analytics , 2016, ArXiv.

[32]  Ian Goodfellow,et al.  Deep Learning with Differential Privacy , 2016, CCS.

[33]  Samy Bengio,et al.  Understanding deep learning requires rethinking generalization , 2016, ICLR.

[34]  Ilya Mironov,et al.  Rényi Differential Privacy , 2017, 2017 IEEE 30th Computer Security Foundations Symposium (CSF).

[35]  Martín Abadi,et al.  Semi-supervised Knowledge Transfer for Deep Learning from Private Training Data , 2016, ICLR.

[36]  Vitaly Shmatikov,et al.  Membership Inference Attacks Against Machine Learning Models , 2016, 2017 IEEE Symposium on Security and Privacy (SP).