Feature weighting by RELIEF based on local hyperplane approximation

Hongmin Cai; Michael Ng

doi:10.1007/978-3-642-30220-6_28

Feature weighting by RELIEF based on local hyperplane approximation

Hongmin Cai^*, Michael Ng

^*Corresponding author for this work

Department of Mathematics

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

7 Citations (Scopus)

Abstract

In this paper, we propose a new feature weighting algorithm through the classical RELIEF framework. The key idea is to estimate the feature weights through local approximation rather than global measurement, as used in previous methods. The weights obtained by our method are more robust to degradation of noisy features, even when the number of dimensions is huge. To demonstrate the performance of our method, we conduct experiments on classification by combining hyperplane KNN model (HKNN) and the proposed feature weight scheme. Empirical study on both synthetic and real-world data sets demonstrate the superior performance of the feature selection for supervised learning, and the effectiveness of our algorithm.

Original language	English
Title of host publication	Advances in Knowledge Discovery and Data Mining - 16th Pacific-Asia Conference, PAKDD 2012, Proceedings
Pages	335-346
Number of pages	12
Edition	PART 2
DOIs	https://doi.org/10.1007/978-3-642-30220-6_28
Publication status	Published - 2012
Event	16th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining, PAKDD 2012 - Kuala Lumpur, Malaysia Duration: 29 May 2012 → 1 Jun 2012

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Number	PART 2
Volume	7301 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	16th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining, PAKDD 2012
Country/Territory	Malaysia
City	Kuala Lumpur
Period	29/05/12 → 1/06/12

Scopus Subject Areas

Theoretical Computer Science
Computer Science(all)

User-Defined Keywords

Classification
Feature weighting
KNN
local hyperplane
RELIEF

Access to Document

10.1007/978-3-642-30220-6_28

Cite this

Cai, H., & Ng, M. (2012). Feature weighting by RELIEF based on local hyperplane approximation. In Advances in Knowledge Discovery and Data Mining - 16th Pacific-Asia Conference, PAKDD 2012, Proceedings (PART 2 ed., pp. 335-346). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7301 LNAI, No. PART 2). https://doi.org/10.1007/978-3-642-30220-6_28

Cai, Hongmin ; Ng, Michael. / Feature weighting by RELIEF based on local hyperplane approximation. Advances in Knowledge Discovery and Data Mining - 16th Pacific-Asia Conference, PAKDD 2012, Proceedings. PART 2. ed. 2012. pp. 335-346 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2).

@inproceedings{c90335d414e94588bed49a6dc175195e,

title = "Feature weighting by RELIEF based on local hyperplane approximation",

abstract = "In this paper, we propose a new feature weighting algorithm through the classical RELIEF framework. The key idea is to estimate the feature weights through local approximation rather than global measurement, as used in previous methods. The weights obtained by our method are more robust to degradation of noisy features, even when the number of dimensions is huge. To demonstrate the performance of our method, we conduct experiments on classification by combining hyperplane KNN model (HKNN) and the proposed feature weight scheme. Empirical study on both synthetic and real-world data sets demonstrate the superior performance of the feature selection for supervised learning, and the effectiveness of our algorithm.",

keywords = "Classification, Feature weighting, KNN, local hyperplane, RELIEF",

author = "Hongmin Cai and Michael Ng",

note = "Copyright: Copyright 2012 Elsevier B.V., All rights reserved.; 16th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining, PAKDD 2012 ; Conference date: 29-05-2012 Through 01-06-2012",

year = "2012",

doi = "10.1007/978-3-642-30220-6_28",

language = "English",

isbn = "9783642302190",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

number = "PART 2",

pages = "335--346",

booktitle = "Advances in Knowledge Discovery and Data Mining - 16th Pacific-Asia Conference, PAKDD 2012, Proceedings",

edition = "PART 2",

}

Cai, H & Ng, M 2012, Feature weighting by RELIEF based on local hyperplane approximation. in Advances in Knowledge Discovery and Data Mining - 16th Pacific-Asia Conference, PAKDD 2012, Proceedings. PART 2 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 2, vol. 7301 LNAI, pp. 335-346, 16th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining, PAKDD 2012, Kuala Lumpur, Malaysia, 29/05/12. https://doi.org/10.1007/978-3-642-30220-6_28

Feature weighting by RELIEF based on local hyperplane approximation. / Cai, Hongmin; Ng, Michael.
Advances in Knowledge Discovery and Data Mining - 16th Pacific-Asia Conference, PAKDD 2012, Proceedings. PART 2. ed. 2012. p. 335-346 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7301 LNAI, No. PART 2).

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

TY - GEN

T1 - Feature weighting by RELIEF based on local hyperplane approximation

AU - Cai, Hongmin

AU - Ng, Michael

PY - 2012

Y1 - 2012

N2 - In this paper, we propose a new feature weighting algorithm through the classical RELIEF framework. The key idea is to estimate the feature weights through local approximation rather than global measurement, as used in previous methods. The weights obtained by our method are more robust to degradation of noisy features, even when the number of dimensions is huge. To demonstrate the performance of our method, we conduct experiments on classification by combining hyperplane KNN model (HKNN) and the proposed feature weight scheme. Empirical study on both synthetic and real-world data sets demonstrate the superior performance of the feature selection for supervised learning, and the effectiveness of our algorithm.

AB - In this paper, we propose a new feature weighting algorithm through the classical RELIEF framework. The key idea is to estimate the feature weights through local approximation rather than global measurement, as used in previous methods. The weights obtained by our method are more robust to degradation of noisy features, even when the number of dimensions is huge. To demonstrate the performance of our method, we conduct experiments on classification by combining hyperplane KNN model (HKNN) and the proposed feature weight scheme. Empirical study on both synthetic and real-world data sets demonstrate the superior performance of the feature selection for supervised learning, and the effectiveness of our algorithm.

KW - Classification

KW - Feature weighting

KW - KNN

KW - local hyperplane

KW - RELIEF

UR - http://www.scopus.com/inward/record.url?scp=84861444374&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-30220-6_28

DO - 10.1007/978-3-642-30220-6_28

M3 - Conference proceeding

AN - SCOPUS:84861444374

SN - 9783642302190

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 335

EP - 346

BT - Advances in Knowledge Discovery and Data Mining - 16th Pacific-Asia Conference, PAKDD 2012, Proceedings

T2 - 16th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining, PAKDD 2012

Y2 - 29 May 2012 through 1 June 2012

ER -

Cai H, Ng M. Feature weighting by RELIEF based on local hyperplane approximation. In Advances in Knowledge Discovery and Data Mining - 16th Pacific-Asia Conference, PAKDD 2012, Proceedings. PART 2 ed. 2012. p. 335-346. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2). doi: 10.1007/978-3-642-30220-6_28

Feature weighting by RELIEF based on local hyperplane approximation

Abstract

Publication series

Conference

Scopus Subject Areas

User-Defined Keywords

Access to Document

Other files and links

Fingerprint

Cite this