HE-Friendly Algorithm for Privacy-Preserving SVM Training

Abstract

Support vector machine (SVM) is one of the most popular machine learning algorithms. It predicts a pre-defined output variable in real-world applications. Machine learning on encrypted data is becoming more and more important to protect both model information and data against various adversaries. While some studies have been proposed on inference or prediction phases, few have been reported on the training phase. Homomorphic encryption (HE) for the arithmetic of approximate numbers scheme enables efficient arithmetic evaluations of encrypted data of real numbers, which encourages to develop privacy-preserving machine learning training algorithm. In this study, we propose an HE-friendly algorithm for the SVM training phase which avoids inefficient operations and numerical instability on an encrypted domain. The inference phase is also implemented on the encrypted domain with fully-homomorphic encryption which enables real-time prediction. Our experiment showed that our HE-friendly algorithm outperformed the state-of-the-art logistic regression classifier with fully homomorphic encryption on toy and real-world datasets. To the best of our knowledge, this study is the first practical algorithm for training an SVM model with fully homomorphic encryption. Therefore, our result supports the development of practical applications of the privacy-preserving SVM model. INDEX TERMS Cryptography, data privacy, fully homomorphic encryption, support vector machine, privacy-preserving training.

Revisiting Multivariate Ring Learning with Errors and Its Applications on Lattice-Based Cryptography

“…Precisely, as in many real scenarios the number of required operations is known beforehand, SHE turns out to be a perfect fit. Furthermore, many optimizations have been incorporated and, consequently, lattice-based SHE/FHE cryptosystems are being progressively adopted by researchers in the field [7,[29][30][31]. In particular, RLWE-based cryptosystems show nowadays the best runtime performance.…”

An Intelligent Detection Method of Personal Privacy Disclosure for Social Networks

“…For event risk value, calculate information leakage risk coefficient according to weight. Reference [9][10][11][12] pointed out that machine learning has been widely applied in the fields of healthcare, cybersecurity, etc. due to its powerful data mining capabilities, where SVM is one of the most popular machine learning algorithms; therefore use SVM algorithm to divide information leakage risk coefficient and get a final evaluation.…”

How to Securely Collaborate on Data: Decentralized Threshold HE and Secure Key Update

“…Since the first construction of HE of Gentry [1], the efficiency has been improved significantly over the decade ( [2]- [14]). Especially, CKKS scheme [15] that supports encrypted approximate arithmetic of real numbers made great strides as it demonstrates the applicability of HE in real use cases ( [16]- [21]). By using HE, organizations can benefit from collaboration without exposing original data, compromising privacy, or concern about re-identification through data integration with other sources.…”