Large margin feature selection for support vector machine

Wei Pan; Peijun Ma; Xiaohong Su

doi:10.4028/www.scientific.net/AMM.274.161

Large margin feature selection for support vector machine

Wei Pan^*
, Peijun Ma
, Xiaohong Su

^*Corresponding author for this work

School of Computer Science and Technology, Harbin Institute of Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

1 Scopus citations

Abstract

Feature selection is an preprocessing step in pattern analysis and machine learning. In this paper, we design a algorithm for feature subset. We present L1-norm regularization technique for sparse feature weight. Margin loss are introduced to evaluate features, and we employs gradient descent to search the optimal solution to maximize margin. The proposed technique is tested on UCI data sets. Compared with four margin based loss functions for SVM, the proposed technique is effective and efficient.

Original language	English
Title of host publication	Mechanical Engineering, Materials Science and Civil Engineering
Publisher	Trans Tech Publications Ltd
Pages	161-164
Number of pages	4
ISBN (Print)	9783037855904
DOIs	https://doi.org/10.4028/www.scientific.net/AMM.274.161
State	Published - 2013
Externally published	Yes
Event	2012 International Conference on Mechanical Engineering, Materials Science and Civil Engineering, ICMEMSCE 2012 - Harbin, China Duration: 18 Aug 2012 → 20 Aug 2012

Publication series

Name	Applied Mechanics and Materials
Volume	274
ISSN (Print)	1660-9336
ISSN (Electronic)	1662-7482

Conference

Conference	2012 International Conference on Mechanical Engineering, Materials Science and Civil Engineering, ICMEMSCE 2012
Country/Territory	China
City	Harbin
Period	18/08/12 → 20/08/12

Keywords

Feature selection
Margin
Support vector machine

Access to Document

10.4028/www.scientific.net/AMM.274.161

Cite this

@inproceedings{2ffbf0fb97f249d49b6503630d8b3842,

title = "Large margin feature selection for support vector machine",

abstract = "Feature selection is an preprocessing step in pattern analysis and machine learning. In this paper, we design a algorithm for feature subset. We present L1-norm regularization technique for sparse feature weight. Margin loss are introduced to evaluate features, and we employs gradient descent to search the optimal solution to maximize margin. The proposed technique is tested on UCI data sets. Compared with four margin based loss functions for SVM, the proposed technique is effective and efficient.",

keywords = "Feature selection, Margin, Support vector machine",

author = "Wei Pan and Peijun Ma and Xiaohong Su",

year = "2013",

doi = "10.4028/www.scientific.net/AMM.274.161",

language = "英语",

isbn = "9783037855904",

series = "Applied Mechanics and Materials",

publisher = "Trans Tech Publications Ltd",

pages = "161--164",

booktitle = "Mechanical Engineering, Materials Science and Civil Engineering",

address = "瑞士",

note = "2012 International Conference on Mechanical Engineering, Materials Science and Civil Engineering, ICMEMSCE 2012 ; Conference date: 18-08-2012 Through 20-08-2012",

}

Pan, W, Ma, P & Su, X 2013, Large margin feature selection for support vector machine. in Mechanical Engineering, Materials Science and Civil Engineering. Applied Mechanics and Materials, vol. 274, Trans Tech Publications Ltd, pp. 161-164, 2012 International Conference on Mechanical Engineering, Materials Science and Civil Engineering, ICMEMSCE 2012, Harbin, China, 18/08/12. https://doi.org/10.4028/www.scientific.net/AMM.274.161

TY - GEN

T1 - Large margin feature selection for support vector machine

AU - Pan, Wei

AU - Ma, Peijun

AU - Su, Xiaohong

PY - 2013

Y1 - 2013

N2 - Feature selection is an preprocessing step in pattern analysis and machine learning. In this paper, we design a algorithm for feature subset. We present L1-norm regularization technique for sparse feature weight. Margin loss are introduced to evaluate features, and we employs gradient descent to search the optimal solution to maximize margin. The proposed technique is tested on UCI data sets. Compared with four margin based loss functions for SVM, the proposed technique is effective and efficient.

AB - Feature selection is an preprocessing step in pattern analysis and machine learning. In this paper, we design a algorithm for feature subset. We present L1-norm regularization technique for sparse feature weight. Margin loss are introduced to evaluate features, and we employs gradient descent to search the optimal solution to maximize margin. The proposed technique is tested on UCI data sets. Compared with four margin based loss functions for SVM, the proposed technique is effective and efficient.

KW - Feature selection

KW - Margin

KW - Support vector machine

UR - https://www.scopus.com/pages/publications/84873860283

U2 - 10.4028/www.scientific.net/AMM.274.161

DO - 10.4028/www.scientific.net/AMM.274.161

M3 - 会议稿件

AN - SCOPUS:84873860283

SN - 9783037855904

T3 - Applied Mechanics and Materials

SP - 161

EP - 164

BT - Mechanical Engineering, Materials Science and Civil Engineering

PB - Trans Tech Publications Ltd

T2 - 2012 International Conference on Mechanical Engineering, Materials Science and Civil Engineering, ICMEMSCE 2012

Y2 - 18 August 2012 through 20 August 2012

ER -

Large margin feature selection for support vector machine

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this