Paper 2020/042
BLAZE: Blazing Fast Privacy-Preserving Machine Learning
Arpita Patra and Ajith Suresh
Abstract
Machine learning tools have illustrated their potential in many significant sectors such as healthcare and finance, to aide in deriving useful inferences. The sensitive and confidential nature of the data, in such sectors, raises natural concerns for the privacy of data. This motivated the area of Privacy-preserving Machine Learning (PPML) where privacy of the data is guaranteed. Typically, ML techniques require large computing power, which leads clients with limited infrastructure to rely on the method of Secure Outsourced Computation (SOC). In SOC setting, the computation is outsourced to a set of specialized and powerful cloud servers and the service is availed on a pay-per-use basis. In this work, we explore PPML techniques in the SOC setting for widely used ML algorithms-- Linear Regression, Logistic Regression, and Neural Networks.
We propose BLAZE, a blazing fast PPML framework in the three server setting tolerating one malicious corruption over a ring (
Note: This article is the full and extended version of an article published in The Network and Distributed System Security Symposium (NDSS) 2020. The article also fixes a small bug, present in one of the protocols of the earlier version.
Metadata
- Available format(s)
-
PDF
- Category
- Cryptographic protocols
- Publication info
- Published elsewhere. Major revision. The Network and Distributed System Security Symposium (NDSS) 2020
- DOI
- 10.14722/ndss.2020.24202
- Keywords
- MPCPPMLPrivacy-preserving Machine LearningMulti-party Computation
- Contact author(s)
-
ajith @ iisc ac in
arpita @ iisc ac in - History
- 2021-01-06: last of 6 revisions
- 2020-01-15: received
- See all versions
- Short URL
- https://ia.cr/2020/042
- License
-
CC BY
BibTeX
@misc{cryptoeprint:2020/042, author = {Arpita Patra and Ajith Suresh}, title = {{BLAZE}: Blazing Fast Privacy-Preserving Machine Learning}, howpublished = {Cryptology {ePrint} Archive, Paper 2020/042}, year = {2020}, doi = {10.14722/ndss.2020.24202}, url = {https://eprint.iacr.org/2020/042} }