On convex stochastic variance reduced gradient for adversarial machine learning

Saikiran Bulusu; Qunwei Li; Pramod K. Varshney

doi:10.1109/GlobalSIP45357.2019.8969103

On convex stochastic variance reduced gradient for adversarial machine learning

Saikiran Bulusu, Qunwei Li, Pramod K. Varshney

Department of Electrical Engineering & Computer Science

Research output: Chapter in Book/Entry/Poem › Conference contribution

3 Scopus citations

Abstract

We study the finite-sum problem in an adversarial setting using stochastic variance reduced gradient (SVRG) optimization in a distributed setting. Here, a fraction of the workers are assumed to be Byzantine that exhibit adversarial behavior by providing arbitrary data. We propose a robust scheme to combat the actions of Byzantine adversaries in this setting, and provide rates of convergence for the convex case. This is the first study of SVRG in an adversarial setting.

Original language	English (US)
Title of host publication	GlobalSIP 2019 - 7th IEEE Global Conference on Signal and Information Processing, Proceedings
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781728127231
DOIs	https://doi.org/10.1109/GlobalSIP45357.2019.8969103
State	Published - Nov 2019
Event	7th IEEE Global Conference on Signal and Information Processing, GlobalSIP 2019 - Ottawa, Canada Duration: Nov 11 2019 → Nov 14 2019

Publication series

Name	GlobalSIP 2019 - 7th IEEE Global Conference on Signal and Information Processing, Proceedings

Conference

Conference	7th IEEE Global Conference on Signal and Information Processing, GlobalSIP 2019
Country/Territory	Canada
City	Ottawa
Period	11/11/19 → 11/14/19

Keywords

Adversarial machine learning
Byzantines
Distributed opti-mization
Stochastic Gradient Descent (SGD)
Stochastic variance reduced gradient (SVRG)

ASJC Scopus subject areas

Information Systems
Information Systems and Management
Artificial Intelligence
Computer Vision and Pattern Recognition
Signal Processing

Access to Document

10.1109/GlobalSIP45357.2019.8969103

Cite this

Bulusu, S., Li, Q., & Varshney, P. K. (2019). On convex stochastic variance reduced gradient for adversarial machine learning. In GlobalSIP 2019 - 7th IEEE Global Conference on Signal and Information Processing, Proceedings Article 8969103 (GlobalSIP 2019 - 7th IEEE Global Conference on Signal and Information Processing, Proceedings). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/GlobalSIP45357.2019.8969103

On convex stochastic variance reduced gradient for adversarial machine learning. / Bulusu, Saikiran; Li, Qunwei; Varshney, Pramod K.
GlobalSIP 2019 - 7th IEEE Global Conference on Signal and Information Processing, Proceedings. Institute of Electrical and Electronics Engineers Inc., 2019. 8969103 (GlobalSIP 2019 - 7th IEEE Global Conference on Signal and Information Processing, Proceedings).

Research output: Chapter in Book/Entry/Poem › Conference contribution

Bulusu, S, Li, Q & Varshney, PK 2019, On convex stochastic variance reduced gradient for adversarial machine learning. in GlobalSIP 2019 - 7th IEEE Global Conference on Signal and Information Processing, Proceedings., 8969103, GlobalSIP 2019 - 7th IEEE Global Conference on Signal and Information Processing, Proceedings, Institute of Electrical and Electronics Engineers Inc., 7th IEEE Global Conference on Signal and Information Processing, GlobalSIP 2019, Ottawa, Canada, 11/11/19. https://doi.org/10.1109/GlobalSIP45357.2019.8969103

Bulusu S, Li Q, Varshney PK. On convex stochastic variance reduced gradient for adversarial machine learning. In GlobalSIP 2019 - 7th IEEE Global Conference on Signal and Information Processing, Proceedings. Institute of Electrical and Electronics Engineers Inc. 2019. 8969103. (GlobalSIP 2019 - 7th IEEE Global Conference on Signal and Information Processing, Proceedings). doi: 10.1109/GlobalSIP45357.2019.8969103

Bulusu, Saikiran ; Li, Qunwei ; Varshney, Pramod K. / On convex stochastic variance reduced gradient for adversarial machine learning. GlobalSIP 2019 - 7th IEEE Global Conference on Signal and Information Processing, Proceedings. Institute of Electrical and Electronics Engineers Inc., 2019. (GlobalSIP 2019 - 7th IEEE Global Conference on Signal and Information Processing, Proceedings).

@inproceedings{d6b4e14dba004742947ba32ab943e790,

title = "On convex stochastic variance reduced gradient for adversarial machine learning",

abstract = "We study the finite-sum problem in an adversarial setting using stochastic variance reduced gradient (SVRG) optimization in a distributed setting. Here, a fraction of the workers are assumed to be Byzantine that exhibit adversarial behavior by providing arbitrary data. We propose a robust scheme to combat the actions of Byzantine adversaries in this setting, and provide rates of convergence for the convex case. This is the first study of SVRG in an adversarial setting.",

keywords = "Adversarial machine learning, Byzantines, Distributed opti-mization, Stochastic Gradient Descent (SGD), Stochastic variance reduced gradient (SVRG)",

author = "Saikiran Bulusu and Qunwei Li and Varshney, {Pramod K.}",

note = "Publisher Copyright: {\textcopyright} 2019 IEEE.; 7th IEEE Global Conference on Signal and Information Processing, GlobalSIP 2019 ; Conference date: 11-11-2019 Through 14-11-2019",

year = "2019",

month = nov,

doi = "10.1109/GlobalSIP45357.2019.8969103",

language = "English (US)",

series = "GlobalSIP 2019 - 7th IEEE Global Conference on Signal and Information Processing, Proceedings",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "GlobalSIP 2019 - 7th IEEE Global Conference on Signal and Information Processing, Proceedings",

}

TY - GEN

T1 - On convex stochastic variance reduced gradient for adversarial machine learning

AU - Bulusu, Saikiran

AU - Li, Qunwei

AU - Varshney, Pramod K.

PY - 2019/11

Y1 - 2019/11

N2 - We study the finite-sum problem in an adversarial setting using stochastic variance reduced gradient (SVRG) optimization in a distributed setting. Here, a fraction of the workers are assumed to be Byzantine that exhibit adversarial behavior by providing arbitrary data. We propose a robust scheme to combat the actions of Byzantine adversaries in this setting, and provide rates of convergence for the convex case. This is the first study of SVRG in an adversarial setting.

AB - We study the finite-sum problem in an adversarial setting using stochastic variance reduced gradient (SVRG) optimization in a distributed setting. Here, a fraction of the workers are assumed to be Byzantine that exhibit adversarial behavior by providing arbitrary data. We propose a robust scheme to combat the actions of Byzantine adversaries in this setting, and provide rates of convergence for the convex case. This is the first study of SVRG in an adversarial setting.

KW - Adversarial machine learning

KW - Byzantines

KW - Distributed opti-mization

KW - Stochastic Gradient Descent (SGD)

KW - Stochastic variance reduced gradient (SVRG)

UR - http://www.scopus.com/inward/record.url?scp=85079269982&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85079269982&partnerID=8YFLogxK

U2 - 10.1109/GlobalSIP45357.2019.8969103

DO - 10.1109/GlobalSIP45357.2019.8969103

M3 - Conference contribution

AN - SCOPUS:85079269982

T3 - GlobalSIP 2019 - 7th IEEE Global Conference on Signal and Information Processing, Proceedings

BT - GlobalSIP 2019 - 7th IEEE Global Conference on Signal and Information Processing, Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 7th IEEE Global Conference on Signal and Information Processing, GlobalSIP 2019

Y2 - 11 November 2019 through 14 November 2019

ER -

On convex stochastic variance reduced gradient for adversarial machine learning

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this