A Differentially Private Stochastic Gradient Descent Algorithm for Multiparty Classification

Arun Rajkumar, Shivani Agarwal
; Proceedings of the Fifteenth International Conference on Artificial Intelligence and Statistics, PMLR 22:933-941, 2012.

Abstract

We consider the problem of developing privacy preserving machine learning algorithms in a distributed multiparty setting. Here different parties own different parts of a data set, and the goal is to learn a classifier from the entire data set without any party revealing any information about the individual data points it owns. Pathak et al (2010) recently proposed a solution to this problem in which each party learns a local classifier from its own data, and a third party then aggregates these classifiers in a privacy-preserving manner using a cryptographic scheme. The generalization performance of their algorithm is sensitive to the number of parties and the relative fractions of data owned by the different parties. In this paper, we describe a new differentially private algorithm for the multiparty setting that uses a stochastic gradient descent based procedure to directly optimize the overall multiparty objective rather than combining classifiers learned from optimizing local objectives. The algorithm achieves a slightly weaker form of differential privacy than that of Pathak et al (2010), but provides improved generalization guarantees that do not depend on the number of parties or the relative sizes of the individual data sets. Experimental results corroborate our theoretical findings.

Cite this Paper


BibTeX
@InProceedings{pmlr-v22-rajkumar12, title = {A Differentially Private Stochastic Gradient Descent Algorithm for Multiparty Classification}, author = {Arun Rajkumar and Shivani Agarwal}, booktitle = {Proceedings of the Fifteenth International Conference on Artificial Intelligence and Statistics}, pages = {933--941}, year = {2012}, editor = {Neil D. Lawrence and Mark Girolami}, volume = {22}, series = {Proceedings of Machine Learning Research}, address = {La Palma, Canary Islands}, month = {21--23 Apr}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v22/rajkumar12/rajkumar12.pdf}, url = {http://proceedings.mlr.press/v22/rajkumar12.html}, abstract = {We consider the problem of developing privacy preserving machine learning algorithms in a distributed multiparty setting. Here different parties own different parts of a data set, and the goal is to learn a classifier from the entire data set without any party revealing any information about the individual data points it owns. Pathak et al (2010) recently proposed a solution to this problem in which each party learns a local classifier from its own data, and a third party then aggregates these classifiers in a privacy-preserving manner using a cryptographic scheme. The generalization performance of their algorithm is sensitive to the number of parties and the relative fractions of data owned by the different parties. In this paper, we describe a new differentially private algorithm for the multiparty setting that uses a stochastic gradient descent based procedure to directly optimize the overall multiparty objective rather than combining classifiers learned from optimizing local objectives. The algorithm achieves a slightly weaker form of differential privacy than that of Pathak et al (2010), but provides improved generalization guarantees that do not depend on the number of parties or the relative sizes of the individual data sets. Experimental results corroborate our theoretical findings.} }
Endnote
%0 Conference Paper %T A Differentially Private Stochastic Gradient Descent Algorithm for Multiparty Classification %A Arun Rajkumar %A Shivani Agarwal %B Proceedings of the Fifteenth International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2012 %E Neil D. Lawrence %E Mark Girolami %F pmlr-v22-rajkumar12 %I PMLR %J Proceedings of Machine Learning Research %P 933--941 %U http://proceedings.mlr.press %V 22 %W PMLR %X We consider the problem of developing privacy preserving machine learning algorithms in a distributed multiparty setting. Here different parties own different parts of a data set, and the goal is to learn a classifier from the entire data set without any party revealing any information about the individual data points it owns. Pathak et al (2010) recently proposed a solution to this problem in which each party learns a local classifier from its own data, and a third party then aggregates these classifiers in a privacy-preserving manner using a cryptographic scheme. The generalization performance of their algorithm is sensitive to the number of parties and the relative fractions of data owned by the different parties. In this paper, we describe a new differentially private algorithm for the multiparty setting that uses a stochastic gradient descent based procedure to directly optimize the overall multiparty objective rather than combining classifiers learned from optimizing local objectives. The algorithm achieves a slightly weaker form of differential privacy than that of Pathak et al (2010), but provides improved generalization guarantees that do not depend on the number of parties or the relative sizes of the individual data sets. Experimental results corroborate our theoretical findings.
RIS
TY - CPAPER TI - A Differentially Private Stochastic Gradient Descent Algorithm for Multiparty Classification AU - Arun Rajkumar AU - Shivani Agarwal BT - Proceedings of the Fifteenth International Conference on Artificial Intelligence and Statistics PY - 2012/03/21 DA - 2012/03/21 ED - Neil D. Lawrence ED - Mark Girolami ID - pmlr-v22-rajkumar12 PB - PMLR SP - 933 DP - PMLR EP - 941 L1 - http://proceedings.mlr.press/v22/rajkumar12/rajkumar12.pdf UR - http://proceedings.mlr.press/v22/rajkumar12.html AB - We consider the problem of developing privacy preserving machine learning algorithms in a distributed multiparty setting. Here different parties own different parts of a data set, and the goal is to learn a classifier from the entire data set without any party revealing any information about the individual data points it owns. Pathak et al (2010) recently proposed a solution to this problem in which each party learns a local classifier from its own data, and a third party then aggregates these classifiers in a privacy-preserving manner using a cryptographic scheme. The generalization performance of their algorithm is sensitive to the number of parties and the relative fractions of data owned by the different parties. In this paper, we describe a new differentially private algorithm for the multiparty setting that uses a stochastic gradient descent based procedure to directly optimize the overall multiparty objective rather than combining classifiers learned from optimizing local objectives. The algorithm achieves a slightly weaker form of differential privacy than that of Pathak et al (2010), but provides improved generalization guarantees that do not depend on the number of parties or the relative sizes of the individual data sets. Experimental results corroborate our theoretical findings. ER -
APA
Rajkumar, A. & Agarwal, S.. (2012). A Differentially Private Stochastic Gradient Descent Algorithm for Multiparty Classification. Proceedings of the Fifteenth International Conference on Artificial Intelligence and Statistics, in PMLR 22:933-941

Related Material