The value of collaboration in convex machine learning with differential privacy

Nan Wu, Farhad Farokhi, David Smith, Mohamed Ali Kaafar

Research output: Chapter in Book/Report/Conference proceedingConference proceeding contributionpeer-review

13 Citations (Scopus)

Abstract

In this paper, we apply machine learning to distributed private data owned by multiple data owners, entities with access to non-overlapping training datasets. We use noisy, differentially-private gradients to minimize the fitness cost of the machine learning model using stochastic gradient descent. We quantify the quality of the trained model, using the fitness cost, as a function of privacy budget and size of the distributed datasets to capture the trade-off between privacy and utility in machine learning. This way, we can predict the outcome of collaboration among privacy-aware data owners prior to executing potentially computationally-expensive machine learning algorithms. Particularly, we show that the difference between the fitness of the trained machine learning model using differentially-private gradient queries and the fitness of the trained machine model in the absence of any privacy concerns is inversely proportional to the size of the training datasets squared and the privacy budget squared. We successfully validate the performance prediction with the actual performance of the proposed privacy-aware learning algorithms, applied to: financial datasets for determining interest rates of loans using regression; and detecting credit card frauds using support vector machines.

Original languageEnglish
Title of host publicationProceedings - 2020 IEEE Symposium on Security and Privacy, SP 2020
Place of PublicationLos Alamitos, CA
PublisherInstitute of Electrical and Electronics Engineers (IEEE)
Pages304-317
Number of pages14
ISBN (Electronic)9781728134970
DOIs
Publication statusPublished - 2020
Event41st IEEE Symposium on Security and Privacy, SP 2020 - San Francisco, United States
Duration: 18 May 202021 May 2020

Publication series

NameProceedings - IEEE Symposium on Security and Privacy
Volume2020-May
ISSN (Print)1081-6011

Conference

Conference41st IEEE Symposium on Security and Privacy, SP 2020
CountryUnited States
CitySan Francisco
Period18/05/2021/05/20

Keywords

  • Differential privacy
  • Machine learning
  • Stochastic gradient algorithm

Fingerprint

Dive into the research topics of 'The value of collaboration in convex machine learning with differential privacy'. Together they form a unique fingerprint.

Cite this