Private learning and sanitization: Pure vs. approximate differential privacy

Amos Beimel; Kobbi Nissim; Uri Stemmer

doi:10.1007/978-3-642-40328-6_26

Private learning and sanitization: Pure vs. approximate differential privacy

Amos Beimel, Kobbi Nissim, Uri Stemmer

Department of Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

80 Scopus citations

Abstract

We compare the sample complexity of private learning and sanitization tasks under pure ε-differential privacy [Dwork, McSherry, Nissim, and Smith TCC 2006] and approximate (ε,δ)-differential privacy [Dwork, Kenthapadi, McSherry, Mironov, and Naor EUROCRYPT 2006]. We show that the sample complexity of these tasks under approximate differential privacy can be significantly lower than that under pure differential privacy.

Original language	English
Title of host publication	Approximation, Randomization, and Combinatorial Optimization
Subtitle of host publication	Algorithms and Techniques - 16th International Workshop, APPROX 2013 and 17th International Workshop, RANDOM 2013, Proceedings
Pages	363-378
Number of pages	16
DOIs	https://doi.org/10.1007/978-3-642-40328-6_26
State	Published - 15 Oct 2013
Event	16th International Workshop on Approximation Algorithms for Combinatorial Optimization Problems, APPROX 2013 and the 17th International Workshop on Randomization and Computation, RANDOM 2013 - Berkeley, CA, United States Duration: 21 Aug 2013 → 23 Aug 2013

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	8096 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	16th International Workshop on Approximation Algorithms for Combinatorial Optimization Problems, APPROX 2013 and the 17th International Workshop on Randomization and Computation, RANDOM 2013
Country/Territory	United States
City	Berkeley, CA
Period	21/08/13 → 23/08/13

Keywords

Differential Privacy
Private Learning
Sanitization

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-642-40328-6_26

Cite this

Beimel, A., Nissim, K., & Stemmer, U. (2013). Private learning and sanitization: Pure vs. approximate differential privacy. In Approximation, Randomization, and Combinatorial Optimization: Algorithms and Techniques - 16th International Workshop, APPROX 2013 and 17th International Workshop, RANDOM 2013, Proceedings (pp. 363-378). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8096 LNCS). https://doi.org/10.1007/978-3-642-40328-6_26

Beimel, Amos ; Nissim, Kobbi ; Stemmer, Uri. / Private learning and sanitization : Pure vs. approximate differential privacy. Approximation, Randomization, and Combinatorial Optimization: Algorithms and Techniques - 16th International Workshop, APPROX 2013 and 17th International Workshop, RANDOM 2013, Proceedings. 2013. pp. 363-378 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{cd274a2e2ec14e81b1dcdf1a4fb0d9d7,

title = "Private learning and sanitization: Pure vs. approximate differential privacy",

abstract = "We compare the sample complexity of private learning and sanitization tasks under pure ε-differential privacy [Dwork, McSherry, Nissim, and Smith TCC 2006] and approximate (ε,δ)-differential privacy [Dwork, Kenthapadi, McSherry, Mironov, and Naor EUROCRYPT 2006]. We show that the sample complexity of these tasks under approximate differential privacy can be significantly lower than that under pure differential privacy.",

keywords = "Differential Privacy, Private Learning, Sanitization",

author = "Amos Beimel and Kobbi Nissim and Uri Stemmer",

note = "Funding Information: Research supported by the Israel Science Foundation (grants No. 938/09 and 2761/12) and by the Frankel Center for Computer Science at Ben-Gurion University. Work done while the second author was a Visiting Scholar at the Harvard Center for Research on Computation and Society (CRCS). Work partially done when the third author was visiting Harvard University supported in part by NSF grant CNS-1237235 and a gift from Google, Inc.; 16th International Workshop on Approximation Algorithms for Combinatorial Optimization Problems, APPROX 2013 and the 17th International Workshop on Randomization and Computation, RANDOM 2013 ; Conference date: 21-08-2013 Through 23-08-2013",

year = "2013",

month = oct,

day = "15",

doi = "10.1007/978-3-642-40328-6_26",

language = "English",

isbn = "9783642403279",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

pages = "363--378",

booktitle = "Approximation, Randomization, and Combinatorial Optimization",

}

Beimel, A, Nissim, K & Stemmer, U 2013, Private learning and sanitization: Pure vs. approximate differential privacy. in Approximation, Randomization, and Combinatorial Optimization: Algorithms and Techniques - 16th International Workshop, APPROX 2013 and 17th International Workshop, RANDOM 2013, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 8096 LNCS, pp. 363-378, 16th International Workshop on Approximation Algorithms for Combinatorial Optimization Problems, APPROX 2013 and the 17th International Workshop on Randomization and Computation, RANDOM 2013, Berkeley, CA, United States, 21/08/13. https://doi.org/10.1007/978-3-642-40328-6_26

Private learning and sanitization: Pure vs. approximate differential privacy. / Beimel, Amos; Nissim, Kobbi; Stemmer, Uri.
Approximation, Randomization, and Combinatorial Optimization: Algorithms and Techniques - 16th International Workshop, APPROX 2013 and 17th International Workshop, RANDOM 2013, Proceedings. 2013. p. 363-378 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8096 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Private learning and sanitization

T2 - 16th International Workshop on Approximation Algorithms for Combinatorial Optimization Problems, APPROX 2013 and the 17th International Workshop on Randomization and Computation, RANDOM 2013

AU - Beimel, Amos

AU - Nissim, Kobbi

AU - Stemmer, Uri

N1 - Funding Information: Research supported by the Israel Science Foundation (grants No. 938/09 and 2761/12) and by the Frankel Center for Computer Science at Ben-Gurion University. Work done while the second author was a Visiting Scholar at the Harvard Center for Research on Computation and Society (CRCS). Work partially done when the third author was visiting Harvard University supported in part by NSF grant CNS-1237235 and a gift from Google, Inc.

PY - 2013/10/15

Y1 - 2013/10/15

N2 - We compare the sample complexity of private learning and sanitization tasks under pure ε-differential privacy [Dwork, McSherry, Nissim, and Smith TCC 2006] and approximate (ε,δ)-differential privacy [Dwork, Kenthapadi, McSherry, Mironov, and Naor EUROCRYPT 2006]. We show that the sample complexity of these tasks under approximate differential privacy can be significantly lower than that under pure differential privacy.

AB - We compare the sample complexity of private learning and sanitization tasks under pure ε-differential privacy [Dwork, McSherry, Nissim, and Smith TCC 2006] and approximate (ε,δ)-differential privacy [Dwork, Kenthapadi, McSherry, Mironov, and Naor EUROCRYPT 2006]. We show that the sample complexity of these tasks under approximate differential privacy can be significantly lower than that under pure differential privacy.

KW - Differential Privacy

KW - Private Learning

KW - Sanitization

UR - http://www.scopus.com/inward/record.url?scp=84885226635&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-40328-6_26

DO - 10.1007/978-3-642-40328-6_26

M3 - Conference contribution

AN - SCOPUS:84885226635

SN - 9783642403279

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 363

EP - 378

BT - Approximation, Randomization, and Combinatorial Optimization

Y2 - 21 August 2013 through 23 August 2013

ER -

Beimel A, Nissim K, Stemmer U. Private learning and sanitization: Pure vs. approximate differential privacy. In Approximation, Randomization, and Combinatorial Optimization: Algorithms and Techniques - 16th International Workshop, APPROX 2013 and 17th International Workshop, RANDOM 2013, Proceedings. 2013. p. 363-378. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-642-40328-6_26

Private learning and sanitization: Pure vs. approximate differential privacy

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this