Distributed and robust support vector machine

Yangwei Liu, Hu Ding, Ziyun Huang, Jinhui Xu

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

1 Citation (Scopus)

Abstract

In this paper, we consider the distributed version of Support Vector Machine (SVM) under the coordinator model, where all input data (i.e., points in ℝd space) of SVM are arbitrarily distributed among k nodes in some network with a coordinator which can communicate with all nodes. We investigate two variants of this problem, with and without outliers. For distributed SVM without outliers, we prove a lower bound on the communication complexity and give a distributed (1 - ε)-approximation algorithm to reach this lower bound, where ε is a user specified small constant. For distributed SVM with outliers, we present a (1 - ε)-approximation algorithm to explicitly remove the influence of outliers. Our algorithm is based on a deterministic distributed top t selection algorithm with communication complexity of O(k log (t)) in the coordinator model. Experimental results on benchmark datasets confirm the theoretical guarantees of our algorithms.

Original languageEnglish (US)
Title of host publication27th International Symposium on Algorithms and Computation, ISAAC 2016
EditorsSeok-Hee Hong
PublisherSchloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH, Dagstuhl Publishing
Pages54.1-54.13
ISBN (Electronic)9783959770262
DOIs
StatePublished - Dec 1 2016
Event27th International Symposium on Algorithms and Computation, ISAAC 2016 - Sydney, Australia
Duration: Dec 12 2016Dec 14 2016

Publication series

NameLeibniz International Proceedings in Informatics, LIPIcs
Volume64
ISSN (Print)1868-8969

Conference

Conference27th International Symposium on Algorithms and Computation, ISAAC 2016
CountryAustralia
CitySydney
Period12/12/1612/14/16

Fingerprint

Support vector machines
Approximation algorithms
Communication

All Science Journal Classification (ASJC) codes

  • Software

Cite this

Liu, Y., Ding, H., Huang, Z., & Xu, J. (2016). Distributed and robust support vector machine. In S-H. Hong (Ed.), 27th International Symposium on Algorithms and Computation, ISAAC 2016 (pp. 54.1-54.13). (Leibniz International Proceedings in Informatics, LIPIcs; Vol. 64). Schloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH, Dagstuhl Publishing. https://doi.org/10.4230/LIPIcs.ISAAC.2016.54
Liu, Yangwei ; Ding, Hu ; Huang, Ziyun ; Xu, Jinhui. / Distributed and robust support vector machine. 27th International Symposium on Algorithms and Computation, ISAAC 2016. editor / Seok-Hee Hong. Schloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH, Dagstuhl Publishing, 2016. pp. 54.1-54.13 (Leibniz International Proceedings in Informatics, LIPIcs).
@inproceedings{ff51806adff843749170a46e0f7b8bac,
title = "Distributed and robust support vector machine",
abstract = "In this paper, we consider the distributed version of Support Vector Machine (SVM) under the coordinator model, where all input data (i.e., points in ℝd space) of SVM are arbitrarily distributed among k nodes in some network with a coordinator which can communicate with all nodes. We investigate two variants of this problem, with and without outliers. For distributed SVM without outliers, we prove a lower bound on the communication complexity and give a distributed (1 - ε)-approximation algorithm to reach this lower bound, where ε is a user specified small constant. For distributed SVM with outliers, we present a (1 - ε)-approximation algorithm to explicitly remove the influence of outliers. Our algorithm is based on a deterministic distributed top t selection algorithm with communication complexity of O(k log (t)) in the coordinator model. Experimental results on benchmark datasets confirm the theoretical guarantees of our algorithms.",
author = "Yangwei Liu and Hu Ding and Ziyun Huang and Jinhui Xu",
year = "2016",
month = "12",
day = "1",
doi = "10.4230/LIPIcs.ISAAC.2016.54",
language = "English (US)",
series = "Leibniz International Proceedings in Informatics, LIPIcs",
publisher = "Schloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH, Dagstuhl Publishing",
pages = "54.1--54.13",
editor = "Seok-Hee Hong",
booktitle = "27th International Symposium on Algorithms and Computation, ISAAC 2016",

}

Liu, Y, Ding, H, Huang, Z & Xu, J 2016, Distributed and robust support vector machine. in S-H Hong (ed.), 27th International Symposium on Algorithms and Computation, ISAAC 2016. Leibniz International Proceedings in Informatics, LIPIcs, vol. 64, Schloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH, Dagstuhl Publishing, pp. 54.1-54.13, 27th International Symposium on Algorithms and Computation, ISAAC 2016, Sydney, Australia, 12/12/16. https://doi.org/10.4230/LIPIcs.ISAAC.2016.54

Distributed and robust support vector machine. / Liu, Yangwei; Ding, Hu; Huang, Ziyun; Xu, Jinhui.

27th International Symposium on Algorithms and Computation, ISAAC 2016. ed. / Seok-Hee Hong. Schloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH, Dagstuhl Publishing, 2016. p. 54.1-54.13 (Leibniz International Proceedings in Informatics, LIPIcs; Vol. 64).

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

TY - GEN

T1 - Distributed and robust support vector machine

AU - Liu, Yangwei

AU - Ding, Hu

AU - Huang, Ziyun

AU - Xu, Jinhui

PY - 2016/12/1

Y1 - 2016/12/1

N2 - In this paper, we consider the distributed version of Support Vector Machine (SVM) under the coordinator model, where all input data (i.e., points in ℝd space) of SVM are arbitrarily distributed among k nodes in some network with a coordinator which can communicate with all nodes. We investigate two variants of this problem, with and without outliers. For distributed SVM without outliers, we prove a lower bound on the communication complexity and give a distributed (1 - ε)-approximation algorithm to reach this lower bound, where ε is a user specified small constant. For distributed SVM with outliers, we present a (1 - ε)-approximation algorithm to explicitly remove the influence of outliers. Our algorithm is based on a deterministic distributed top t selection algorithm with communication complexity of O(k log (t)) in the coordinator model. Experimental results on benchmark datasets confirm the theoretical guarantees of our algorithms.

AB - In this paper, we consider the distributed version of Support Vector Machine (SVM) under the coordinator model, where all input data (i.e., points in ℝd space) of SVM are arbitrarily distributed among k nodes in some network with a coordinator which can communicate with all nodes. We investigate two variants of this problem, with and without outliers. For distributed SVM without outliers, we prove a lower bound on the communication complexity and give a distributed (1 - ε)-approximation algorithm to reach this lower bound, where ε is a user specified small constant. For distributed SVM with outliers, we present a (1 - ε)-approximation algorithm to explicitly remove the influence of outliers. Our algorithm is based on a deterministic distributed top t selection algorithm with communication complexity of O(k log (t)) in the coordinator model. Experimental results on benchmark datasets confirm the theoretical guarantees of our algorithms.

UR - http://www.scopus.com/inward/record.url?scp=85010735785&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85010735785&partnerID=8YFLogxK

U2 - 10.4230/LIPIcs.ISAAC.2016.54

DO - 10.4230/LIPIcs.ISAAC.2016.54

M3 - Conference contribution

AN - SCOPUS:85010735785

T3 - Leibniz International Proceedings in Informatics, LIPIcs

SP - 54.1-54.13

BT - 27th International Symposium on Algorithms and Computation, ISAAC 2016

A2 - Hong, Seok-Hee

PB - Schloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH, Dagstuhl Publishing

ER -

Liu Y, Ding H, Huang Z, Xu J. Distributed and robust support vector machine. In Hong S-H, editor, 27th International Symposium on Algorithms and Computation, ISAAC 2016. Schloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH, Dagstuhl Publishing. 2016. p. 54.1-54.13. (Leibniz International Proceedings in Informatics, LIPIcs). https://doi.org/10.4230/LIPIcs.ISAAC.2016.54