Linear semi-supervised dimensionality reduction with pairwise constraint for multiple subclasses

Bin Tong, Weifeng Jia, Yanli Ji, Einoshin Suzuki

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

We propose a new method, called Subclass-oriented Dimensionality Reduction with Pairwise Constraints (SODRPaC), for dimensionality reduction. In a high dimensional space, it is common that a group of data points with one class may scatter in several different groups. Current linear semi-supervised dimensionality reduction methods would fail to achieve fair performances, as they assume two data points linked by a must-link constraint are close each other, while they are likely to be located in different groups. Inspired by the above observation, we classify the must-link constraint into two categories, which are the inter-subclass must-link constraint and the intra-subclass must-link constraint, respectively. We carefully generate cannot-link constraints by using must-link constraints, and then propose a new discriminant criterion by employing the cannot-link constraints and the compactness of shared nearest neighbors. The manifold regularization is also incorporated in our dimensionality reduction framework. Extensive experiments on both synthetic and practical data sets illustrate the effectiveness of our method.

Original languageEnglish
Pages (from-to)812-820
Number of pages9
JournalIEICE Transactions on Information and Systems
VolumeE95-D
Issue number3
DOIs
Publication statusPublished - Jan 1 2012

Fingerprint

Experiments

All Science Journal Classification (ASJC) codes

  • Software
  • Hardware and Architecture
  • Computer Vision and Pattern Recognition
  • Electrical and Electronic Engineering
  • Artificial Intelligence

Cite this

Linear semi-supervised dimensionality reduction with pairwise constraint for multiple subclasses. / Tong, Bin; Jia, Weifeng; Ji, Yanli; Suzuki, Einoshin.

In: IEICE Transactions on Information and Systems, Vol. E95-D, No. 3, 01.01.2012, p. 812-820.

Research output: Contribution to journalArticle

@article{0e81d02e6221456581c5da3b2a2c5ca3,
title = "Linear semi-supervised dimensionality reduction with pairwise constraint for multiple subclasses",
abstract = "We propose a new method, called Subclass-oriented Dimensionality Reduction with Pairwise Constraints (SODRPaC), for dimensionality reduction. In a high dimensional space, it is common that a group of data points with one class may scatter in several different groups. Current linear semi-supervised dimensionality reduction methods would fail to achieve fair performances, as they assume two data points linked by a must-link constraint are close each other, while they are likely to be located in different groups. Inspired by the above observation, we classify the must-link constraint into two categories, which are the inter-subclass must-link constraint and the intra-subclass must-link constraint, respectively. We carefully generate cannot-link constraints by using must-link constraints, and then propose a new discriminant criterion by employing the cannot-link constraints and the compactness of shared nearest neighbors. The manifold regularization is also incorporated in our dimensionality reduction framework. Extensive experiments on both synthetic and practical data sets illustrate the effectiveness of our method.",
author = "Bin Tong and Weifeng Jia and Yanli Ji and Einoshin Suzuki",
year = "2012",
month = "1",
day = "1",
doi = "10.1587/transinf.E95.D.812",
language = "English",
volume = "E95-D",
pages = "812--820",
journal = "IEICE Transactions on Information and Systems",
issn = "0916-8532",
publisher = "一般社団法人電子情報通信学会",
number = "3",

}

TY - JOUR

T1 - Linear semi-supervised dimensionality reduction with pairwise constraint for multiple subclasses

AU - Tong, Bin

AU - Jia, Weifeng

AU - Ji, Yanli

AU - Suzuki, Einoshin

PY - 2012/1/1

Y1 - 2012/1/1

N2 - We propose a new method, called Subclass-oriented Dimensionality Reduction with Pairwise Constraints (SODRPaC), for dimensionality reduction. In a high dimensional space, it is common that a group of data points with one class may scatter in several different groups. Current linear semi-supervised dimensionality reduction methods would fail to achieve fair performances, as they assume two data points linked by a must-link constraint are close each other, while they are likely to be located in different groups. Inspired by the above observation, we classify the must-link constraint into two categories, which are the inter-subclass must-link constraint and the intra-subclass must-link constraint, respectively. We carefully generate cannot-link constraints by using must-link constraints, and then propose a new discriminant criterion by employing the cannot-link constraints and the compactness of shared nearest neighbors. The manifold regularization is also incorporated in our dimensionality reduction framework. Extensive experiments on both synthetic and practical data sets illustrate the effectiveness of our method.

AB - We propose a new method, called Subclass-oriented Dimensionality Reduction with Pairwise Constraints (SODRPaC), for dimensionality reduction. In a high dimensional space, it is common that a group of data points with one class may scatter in several different groups. Current linear semi-supervised dimensionality reduction methods would fail to achieve fair performances, as they assume two data points linked by a must-link constraint are close each other, while they are likely to be located in different groups. Inspired by the above observation, we classify the must-link constraint into two categories, which are the inter-subclass must-link constraint and the intra-subclass must-link constraint, respectively. We carefully generate cannot-link constraints by using must-link constraints, and then propose a new discriminant criterion by employing the cannot-link constraints and the compactness of shared nearest neighbors. The manifold regularization is also incorporated in our dimensionality reduction framework. Extensive experiments on both synthetic and practical data sets illustrate the effectiveness of our method.

UR - http://www.scopus.com/inward/record.url?scp=84857873028&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84857873028&partnerID=8YFLogxK

U2 - 10.1587/transinf.E95.D.812

DO - 10.1587/transinf.E95.D.812

M3 - Article

AN - SCOPUS:84857873028

VL - E95-D

SP - 812

EP - 820

JO - IEICE Transactions on Information and Systems

JF - IEICE Transactions on Information and Systems

SN - 0916-8532

IS - 3

ER -