PD-ML-Lite: Private Distributed Machine Learning from Lightweight Cryptography

Maksim Tsikhanovich*, Malik Magdon-Ismail, Muhammad Ishaq, Vassilis Zikas

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Privacy arises to a major issue in distributed learning. Current approaches that do not use a trusted external authority either reduce the accuracy of the learning algorithm (e.g., by adding noise), or incur a high performance penalty. We propose a methodology for private distributed ML from light-weight cryptography (in short, PD-ML-Lite). We apply our methodology to two major ML algorithms, namely non-negative matrix factorization (NMF) and singular value decomposition (SVD). Our protocols are communication optimal, achieve the same accuracy as their non-private counterparts, and satisfy a notion of privacy—which we define—that is both intuitive and measurable. We use light cryptographic tools (multi-party secure sum and normed secure sum) to build learning algorithms rather than wrap complex learning algorithms in a heavy multi-party computation (MPC) framework. We showcase our algorithms’ utility and privacy for NMF on topic modeling and recommender systems, and for SVD on principal component regression, and low rank approximation.

Original languageEnglish
Title of host publicationInformation Security
Subtitle of host publication22nd International Conference, ISC 2019, Proceedings
EditorsZhiqiang Lin, Charalampos Papamanthou, Michalis Polychronakis
Place of PublicationCham
PublisherSpringer
Pages149-167
Number of pages19
ISBN (Electronic)978-3-030-30215-3
ISBN (Print)978-3-030-30214-6
DOIs
Publication statusPublished - 2 Sep 2019
Event22nd International Conference on Information Security: ISC 2019 - New York City, United States
Duration: 16 Sep 201918 Sep 2019

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
PublisherSpringer, Cham
Volume11723
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference22nd International Conference on Information Security
CountryUnited States
CityNew York City
Period16/09/1918/09/19

Fingerprint Dive into the research topics of 'PD-ML-Lite: Private Distributed Machine Learning from Lightweight Cryptography'. Together they form a unique fingerprint.

Cite this