A Bitwise Logistic Regression Using Binary Approximation and Real Number Division in Homomorphic Encryption Scheme

Joon Soo Yoo, Jeong Hwan Hwang, Baek Kyung Song, Ji Won Yoon

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

Abstract

Homomorphic Encryption (HE) is considered to be one of the most promising solutions to maintain secure data outsourcing because the user’s query is processed under encrypted state. Accordingly, many of existing literature related to HE utilizes additive and multiplicative property of HE to facilitate logistic regression which requires high precision for prediction. In consequence, they inevitably transform or approximate nonlinear function of the logistic regression to adjust to their scheme using simple polynomial approximation algorithms such as Taylor expansion. However, such an approximation can be used only in limited applications because they cause unwanted error in results if the function is highly nonlinear. In response, we propose a different approximation approach to constructing the highly accurate logistic regression for HE using binary approximation. Our novel approach originates from bitwise operations on encrypted bits to designing (1) real number representation, (2) division and (3) exponential function. The result of our experiment shows that our approach can be more generally applied and accuracy-guaranteed than the current literature.

Original languageEnglish
Title of host publicationInformation Security Practice and Experience - 15th International Conference, ISPEC 2019, Proceedings
EditorsSwee-Huay Heng, Javier Lopez
PublisherSpringer
Pages20-40
Number of pages21
ISBN (Print)9783030343385
DOIs
Publication statusPublished - 2019 Jan 1
Event15th International Conference on Information Security Practice and Experience, ISPEC 2019 - Kuala Lumpur, Malaysia
Duration: 2019 Nov 262019 Nov 28

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11879 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference15th International Conference on Information Security Practice and Experience, ISPEC 2019
CountryMalaysia
CityKuala Lumpur
Period19/11/2619/11/28

Fingerprint

Homomorphic Encryption
Logistic Regression
Cryptography
Logistics
Division
Binary
Approximation
Polynomial approximation
Outsourcing
Exponential functions
Taylor Expansion
Polynomial Algorithm
Approximation algorithms
Polynomial Approximation
Nonlinear Function
Approximation Algorithms
Multiplicative
Query
Transform
Prediction

Keywords

  • Bitwise operation
  • Homomorphic Encryption
  • Sigmoid

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Yoo, J. S., Hwang, J. H., Song, B. K., & Yoon, J. W. (2019). A Bitwise Logistic Regression Using Binary Approximation and Real Number Division in Homomorphic Encryption Scheme. In S-H. Heng, & J. Lopez (Eds.), Information Security Practice and Experience - 15th International Conference, ISPEC 2019, Proceedings (pp. 20-40). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11879 LNCS). Springer. https://doi.org/10.1007/978-3-030-34339-2_2

A Bitwise Logistic Regression Using Binary Approximation and Real Number Division in Homomorphic Encryption Scheme. / Yoo, Joon Soo; Hwang, Jeong Hwan; Song, Baek Kyung; Yoon, Ji Won.

Information Security Practice and Experience - 15th International Conference, ISPEC 2019, Proceedings. ed. / Swee-Huay Heng; Javier Lopez. Springer, 2019. p. 20-40 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11879 LNCS).

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

Yoo, JS, Hwang, JH, Song, BK & Yoon, JW 2019, A Bitwise Logistic Regression Using Binary Approximation and Real Number Division in Homomorphic Encryption Scheme. in S-H Heng & J Lopez (eds), Information Security Practice and Experience - 15th International Conference, ISPEC 2019, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11879 LNCS, Springer, pp. 20-40, 15th International Conference on Information Security Practice and Experience, ISPEC 2019, Kuala Lumpur, Malaysia, 19/11/26. https://doi.org/10.1007/978-3-030-34339-2_2
Yoo JS, Hwang JH, Song BK, Yoon JW. A Bitwise Logistic Regression Using Binary Approximation and Real Number Division in Homomorphic Encryption Scheme. In Heng S-H, Lopez J, editors, Information Security Practice and Experience - 15th International Conference, ISPEC 2019, Proceedings. Springer. 2019. p. 20-40. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-030-34339-2_2
Yoo, Joon Soo ; Hwang, Jeong Hwan ; Song, Baek Kyung ; Yoon, Ji Won. / A Bitwise Logistic Regression Using Binary Approximation and Real Number Division in Homomorphic Encryption Scheme. Information Security Practice and Experience - 15th International Conference, ISPEC 2019, Proceedings. editor / Swee-Huay Heng ; Javier Lopez. Springer, 2019. pp. 20-40 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{9cb27c32699241178637819edc60b9a5,
title = "A Bitwise Logistic Regression Using Binary Approximation and Real Number Division in Homomorphic Encryption Scheme",
abstract = "Homomorphic Encryption (HE) is considered to be one of the most promising solutions to maintain secure data outsourcing because the user’s query is processed under encrypted state. Accordingly, many of existing literature related to HE utilizes additive and multiplicative property of HE to facilitate logistic regression which requires high precision for prediction. In consequence, they inevitably transform or approximate nonlinear function of the logistic regression to adjust to their scheme using simple polynomial approximation algorithms such as Taylor expansion. However, such an approximation can be used only in limited applications because they cause unwanted error in results if the function is highly nonlinear. In response, we propose a different approximation approach to constructing the highly accurate logistic regression for HE using binary approximation. Our novel approach originates from bitwise operations on encrypted bits to designing (1) real number representation, (2) division and (3) exponential function. The result of our experiment shows that our approach can be more generally applied and accuracy-guaranteed than the current literature.",
keywords = "Bitwise operation, Homomorphic Encryption, Sigmoid",
author = "Yoo, {Joon Soo} and Hwang, {Jeong Hwan} and Song, {Baek Kyung} and Yoon, {Ji Won}",
year = "2019",
month = "1",
day = "1",
doi = "10.1007/978-3-030-34339-2_2",
language = "English",
isbn = "9783030343385",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer",
pages = "20--40",
editor = "Swee-Huay Heng and Javier Lopez",
booktitle = "Information Security Practice and Experience - 15th International Conference, ISPEC 2019, Proceedings",

}

TY - GEN

T1 - A Bitwise Logistic Regression Using Binary Approximation and Real Number Division in Homomorphic Encryption Scheme

AU - Yoo, Joon Soo

AU - Hwang, Jeong Hwan

AU - Song, Baek Kyung

AU - Yoon, Ji Won

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Homomorphic Encryption (HE) is considered to be one of the most promising solutions to maintain secure data outsourcing because the user’s query is processed under encrypted state. Accordingly, many of existing literature related to HE utilizes additive and multiplicative property of HE to facilitate logistic regression which requires high precision for prediction. In consequence, they inevitably transform or approximate nonlinear function of the logistic regression to adjust to their scheme using simple polynomial approximation algorithms such as Taylor expansion. However, such an approximation can be used only in limited applications because they cause unwanted error in results if the function is highly nonlinear. In response, we propose a different approximation approach to constructing the highly accurate logistic regression for HE using binary approximation. Our novel approach originates from bitwise operations on encrypted bits to designing (1) real number representation, (2) division and (3) exponential function. The result of our experiment shows that our approach can be more generally applied and accuracy-guaranteed than the current literature.

AB - Homomorphic Encryption (HE) is considered to be one of the most promising solutions to maintain secure data outsourcing because the user’s query is processed under encrypted state. Accordingly, many of existing literature related to HE utilizes additive and multiplicative property of HE to facilitate logistic regression which requires high precision for prediction. In consequence, they inevitably transform or approximate nonlinear function of the logistic regression to adjust to their scheme using simple polynomial approximation algorithms such as Taylor expansion. However, such an approximation can be used only in limited applications because they cause unwanted error in results if the function is highly nonlinear. In response, we propose a different approximation approach to constructing the highly accurate logistic regression for HE using binary approximation. Our novel approach originates from bitwise operations on encrypted bits to designing (1) real number representation, (2) division and (3) exponential function. The result of our experiment shows that our approach can be more generally applied and accuracy-guaranteed than the current literature.

KW - Bitwise operation

KW - Homomorphic Encryption

KW - Sigmoid

UR - http://www.scopus.com/inward/record.url?scp=85076685608&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85076685608&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-34339-2_2

DO - 10.1007/978-3-030-34339-2_2

M3 - Conference contribution

AN - SCOPUS:85076685608

SN - 9783030343385

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 20

EP - 40

BT - Information Security Practice and Experience - 15th International Conference, ISPEC 2019, Proceedings

A2 - Heng, Swee-Huay

A2 - Lopez, Javier

PB - Springer

ER -