Population-guided large margin classifier for high-dimension low-sample-size problems

Qingbo Yin, Ehsan Adeli, Liran Shen, Dinggang Shen

Research output: Contribution to journalArticle

Abstract

In this paper, we propose a novel linear binary classifier, denoted by population-guided large margin classifier (PGLMC), applicable to any sorts of data, including high-dimensional low-sample-size (HDLSS). PGLMC is conceived with a projecting direction w given by the comprehensive consideration of local structural information of the hyperplane and the statistics of the training samples. Our proposed model has several advantages compared to those widely used approaches. First, it isn't sensitive to the intercept term b. Second, it operates well with imbalanced data. Third, it is relatively simple to be implemented based on Quadratic Programming. Fourth, it is robust to the model specification for various real applications. The theoretical properties of PGLMC are proven. We conduct a series of evaluations on the simulated and five realworld benchmark data sets, including DNA classification, medical image analysis and face recognition. PGLMC outperforms the state-of-theart classification methods in most cases, or obtains comparable results.

Original languageEnglish
Article number107030
JournalPattern Recognition
Volume97
DOIs
Publication statusPublished - 2020 Jan 1

Fingerprint

Classifiers
Quadratic programming
Face recognition
Image analysis
DNA
Statistics
Specifications

Keywords

  • Binary linear classifier
  • Data piling
  • High-dimension lowsample-size
  • Hyperplane
  • Large margin classification
  • Local structure information

ASJC Scopus subject areas

  • Software
  • Signal Processing
  • Computer Vision and Pattern Recognition
  • Artificial Intelligence

Cite this

Population-guided large margin classifier for high-dimension low-sample-size problems. / Yin, Qingbo; Adeli, Ehsan; Shen, Liran; Shen, Dinggang.

In: Pattern Recognition, Vol. 97, 107030, 01.01.2020.

Research output: Contribution to journalArticle

@article{16dbf70a52594634843c3f5aa714575e,
title = "Population-guided large margin classifier for high-dimension low-sample-size problems",
abstract = "In this paper, we propose a novel linear binary classifier, denoted by population-guided large margin classifier (PGLMC), applicable to any sorts of data, including high-dimensional low-sample-size (HDLSS). PGLMC is conceived with a projecting direction w given by the comprehensive consideration of local structural information of the hyperplane and the statistics of the training samples. Our proposed model has several advantages compared to those widely used approaches. First, it isn't sensitive to the intercept term b. Second, it operates well with imbalanced data. Third, it is relatively simple to be implemented based on Quadratic Programming. Fourth, it is robust to the model specification for various real applications. The theoretical properties of PGLMC are proven. We conduct a series of evaluations on the simulated and five realworld benchmark data sets, including DNA classification, medical image analysis and face recognition. PGLMC outperforms the state-of-theart classification methods in most cases, or obtains comparable results.",
keywords = "Binary linear classifier, Data piling, High-dimension lowsample-size, Hyperplane, Large margin classification, Local structure information",
author = "Qingbo Yin and Ehsan Adeli and Liran Shen and Dinggang Shen",
year = "2020",
month = "1",
day = "1",
doi = "10.1016/j.patcog.2019.107030",
language = "English",
volume = "97",
journal = "Pattern Recognition",
issn = "0031-3203",
publisher = "Elsevier Limited",

}

TY - JOUR

T1 - Population-guided large margin classifier for high-dimension low-sample-size problems

AU - Yin, Qingbo

AU - Adeli, Ehsan

AU - Shen, Liran

AU - Shen, Dinggang

PY - 2020/1/1

Y1 - 2020/1/1

N2 - In this paper, we propose a novel linear binary classifier, denoted by population-guided large margin classifier (PGLMC), applicable to any sorts of data, including high-dimensional low-sample-size (HDLSS). PGLMC is conceived with a projecting direction w given by the comprehensive consideration of local structural information of the hyperplane and the statistics of the training samples. Our proposed model has several advantages compared to those widely used approaches. First, it isn't sensitive to the intercept term b. Second, it operates well with imbalanced data. Third, it is relatively simple to be implemented based on Quadratic Programming. Fourth, it is robust to the model specification for various real applications. The theoretical properties of PGLMC are proven. We conduct a series of evaluations on the simulated and five realworld benchmark data sets, including DNA classification, medical image analysis and face recognition. PGLMC outperforms the state-of-theart classification methods in most cases, or obtains comparable results.

AB - In this paper, we propose a novel linear binary classifier, denoted by population-guided large margin classifier (PGLMC), applicable to any sorts of data, including high-dimensional low-sample-size (HDLSS). PGLMC is conceived with a projecting direction w given by the comprehensive consideration of local structural information of the hyperplane and the statistics of the training samples. Our proposed model has several advantages compared to those widely used approaches. First, it isn't sensitive to the intercept term b. Second, it operates well with imbalanced data. Third, it is relatively simple to be implemented based on Quadratic Programming. Fourth, it is robust to the model specification for various real applications. The theoretical properties of PGLMC are proven. We conduct a series of evaluations on the simulated and five realworld benchmark data sets, including DNA classification, medical image analysis and face recognition. PGLMC outperforms the state-of-theart classification methods in most cases, or obtains comparable results.

KW - Binary linear classifier

KW - Data piling

KW - High-dimension lowsample-size

KW - Hyperplane

KW - Large margin classification

KW - Local structure information

UR - http://www.scopus.com/inward/record.url?scp=85071843074&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85071843074&partnerID=8YFLogxK

U2 - 10.1016/j.patcog.2019.107030

DO - 10.1016/j.patcog.2019.107030

M3 - Article

AN - SCOPUS:85071843074

VL - 97

JO - Pattern Recognition

JF - Pattern Recognition

SN - 0031-3203

M1 - 107030

ER -