Abstract
Learning is a task that generalizes many of the analyses that are applied to collections of data, and in particular, collections of sensitive individual information. Hence, it is natural to ask what can be learned while preserving individual privacy. [Kasiviswanathan, Lee, Nissim, Raskhodnikova, and Smith; FOCS 2008] initiated such a discussion. They formalized the notion of private learning, as a combination of PAC learning and differential privacy, and investigated what concept classes can be learned privately. Somewhat surprisingly, they showed that, ignoring time complexity, every PAC learning task could be performed privately with polynomially many samples, and in many natural cases this could even be done in polynomial time.
While these results seem to equate non-private and private learning, there is still a significant gap: the sample complexity of (non-private) PAC learning is crisply characterized in terms of the VC-dimension of the concept class, whereas this relationship is lost in the constructions of private learners, which exhibit, generally, a higher sample complexity.
Looking into this gap, we examine several private learning tasks and give tight bounds on their sample complexity. In particular, we show strong separations between sample complexities of proper and improper private learners (such separation does not exist for non-private learners), and between sample complexities of efficient and inefficient proper private learners. Our results show that VC-dimension is not the right measure for characterizing the sample complexity of proper private learning.
We also examine the task of private data release (as initiated by [Blum, Ligett, and Roth; STOC 2008]), and give new lower bounds on the sample complexity. Our results show that the logarithmic dependence on size of the instance space is essential for private data release.
The original version of this chapter was revised: The copyright line was incorrect. This has been corrected. The Erratum to this chapter is available at DOI: 10.1007/978-3-642-11799-2_36
Chapter PDF
References
Beimel, A., Kasiviswanathan, S., Nissim, K.: Bounds on the Sample Complexity for Private Learning and Private Data Release (Full version) (2009)
Blum, A., Dwork, C., McSherry, F., Nissim, K.: Practical privacy: The SuLQ framework. In: PODS, pp. 128–138. ACM, New York (2005)
Blum, A., Ligett, K., Roth, A.: A learning theory approach to non-interactive database privacy. In: STOC, pp. 609–618. ACM, New York (2008)
Blum, A., Ligett, K., Roth, A.: Private communication (2008)
Blumer, A., Ehrenfeucht, A., Haussler, D., Warmuth, M.K.: Learnability and the Vapnik-Chervonenkis dimension. Journal of the Association for Computing Machinery 36(4), 929–965 (1989)
Dwork, C.: The differential privacy frontier (extended abstract). In: Reingold, O. (ed.) TCC 2009. LNCS, vol. 5444, pp. 496–502. Springer, Heidelberg (2009)
Dwork, C., McSherry, F., Nissim, K., Smith, A.: Calibrating noise to sensitivity in private data analysis. In: Halevi, S., Rabin, T. (eds.) TCC 2006. LNCS, vol. 3876, pp. 265–284. Springer, Heidelberg (2006)
Dwork, C., Naor, M., Reingold, O., Rothblum, G., Vadhan, S.: On the complexity of differentially private data release. In: STOC, pp. 381–390. ACM, New York (2009)
Ehrenfeucht, A., Haussler, D., Kearns, M.J., Valiant, L.G.: A general lower bound on the number of examples needed for learning. Inf. Comput. 82(3), 247–261 (1989)
Kasiviswanathan, S.P., Lee, H.K., Nissim, K., Raskhodnikova, S., Smith, A.: What can we learn privately? In: FOCS, pp. 531–540. IEEE Computer Society, Los Alamitos (2008)
Kasiviswanathan, S.P., Smith, A.: A note on differential privacy: Defining resistance to arbitrary side information. CoRR, arXiv:0803.39461 [cs.CR] (2008)
Kearns, M.J.: Efficient noise-tolerant learning from statistical queries. Journal of the ACM 45(6), 983–1006 (1998); Preliminary version in Proceedings of STOC 1993
Kearns, M.J., Vazirani, U.V.: An Introduction to Computational Learning Theory. MIT Press, Cambridge (1994)
McSherry, F., Talwar, K.: Mechanism design via differential privacy. In: FOCS, pp. 94–103. IEEE, Los Alamitos (2007)
Mishra, N., Sandler, M.: Privacy via pseudorandom sketches. In: PODS, pp. 143–152. ACM, New York (2006)
Pitt, L., Valiant, L.G.: Computational limitations on learning from examples. Journal of the ACM 35(4), 965–984 (1988)
Valiant, L.G.: A theory of the learnable. Communications of the ACM 27, 1134–1142 (1984)
Vapnik, V.N., Chervonenkis, A.Y.: On the uniform convergence of relative frequencies of events to their probabilities. Theory of Probability and its Applications 16, 264 (1971)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2010 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Beimel, A., Kasiviswanathan, S.P., Nissim, K. (2010). Bounds on the Sample Complexity for Private Learning and Private Data Release. In: Micciancio, D. (eds) Theory of Cryptography. TCC 2010. Lecture Notes in Computer Science, vol 5978. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-11799-2_26
Download citation
DOI: https://doi.org/10.1007/978-3-642-11799-2_26
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-11798-5
Online ISBN: 978-3-642-11799-2
eBook Packages: Computer ScienceComputer Science (R0)