Spectral Energy Minimization for Semi-supervised Learning

Chun-hung Li; Zhi-li Wu

doi:10.1007/978-3-540-24775-3_4

Spectral Energy Minimization for Semi-supervised Learning

Chun-hung Li, Zhi-li Wu

Department of Computer Science

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

1 Citation (Scopus)

Abstract

Data mining problems often involve a large amount of unlabeled data and there is often very limited known information on the dataset. In such scenario, semi-supervised learning can often improve classification performance by utilizing unlabeled data for learning. In this paper, we proposed a novel approach to semi-supervised learning as as an optimization of both the classification energy and cluster compactness energy in the unlabeled dataset. The resulting integer programming problem is relaxed by a semi-definite relaxation where efficient solution can be obtained. Furthermore, the spectral graph methods provide improved energy minimization via the incorporation of additional criteria. Results on UCI datasets show promising results.

Original language	English
Title of host publication	Advances in Knowledge Discovery and Data Mining
Subtitle of host publication	8th Pacific-Asia Conference, PAKDD 2004, Sydney, Australia, May 26-28, 2004, Proceedings
Editors	Honghua Dai, Ramakrishnan Srikant, Chengqi Zhang
Publisher	Springer Berlin Heidelberg
Pages	13-21
Number of pages	9
Edition	1st
ISBN (Electronic)	9783540247753
ISBN (Print)	9783540220640
DOIs	https://doi.org/10.1007/978-3-540-24775-3_4
Publication status	Published - 22 Apr 2004
Event	8th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining, PAKDD 2004 - Sydney, Australia Duration: 26 May 2004 → 28 May 2004 https://link.springer.com/book/10.1007/b97861 (Conference proceedings)

Publication series

Name	Lecture Notes in Computer Science
Volume	3056
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Name	Lecture Notes in Artificial Intelligence

Name	PAKDD: Pacific-Asia Conference on Knowledge Discovery and Data Mining

Conference

Conference	8th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining, PAKDD 2004
Abbreviated title	PAKDD 2004
Country/Territory	Australia
City	Sydney
Period	26/05/04 → 28/05/04
Internet address	https://link.springer.com/book/10.1007/b97861 (Conference proceedings)

User-Defined Keywords

Spectral Method
Label Data
Unlabeled Data
Data Mining Problem
Transductive Support Vector Machine

Access to Document

10.1007/978-3-540-24775-3_4

Cite this

Li, C., & Wu, Z. (2004). Spectral Energy Minimization for Semi-supervised Learning. In H. Dai, R. Srikant, & C. Zhang (Eds.), Advances in Knowledge Discovery and Data Mining: 8th Pacific-Asia Conference, PAKDD 2004, Sydney, Australia, May 26-28, 2004, Proceedings (1st ed., pp. 13-21). (Lecture Notes in Computer Science; Vol. 3056), (Lecture Notes in Artificial Intelligence), (PAKDD: Pacific-Asia Conference on Knowledge Discovery and Data Mining). Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-540-24775-3_4

Li, Chun-hung ; Wu, Zhi-li. / Spectral Energy Minimization for Semi-supervised Learning. Advances in Knowledge Discovery and Data Mining: 8th Pacific-Asia Conference, PAKDD 2004, Sydney, Australia, May 26-28, 2004, Proceedings. editor / Honghua Dai ; Ramakrishnan Srikant ; Chengqi Zhang. 1st. ed. Springer Berlin Heidelberg, 2004. pp. 13-21 (Lecture Notes in Computer Science). (Lecture Notes in Artificial Intelligence). (PAKDD: Pacific-Asia Conference on Knowledge Discovery and Data Mining).

@inproceedings{9aa84a2db3a44f92a77974b9464ad8bf,

title = "Spectral Energy Minimization for Semi-supervised Learning",

abstract = "Data mining problems often involve a large amount of unlabeled data and there is often very limited known information on the dataset. In such scenario, semi-supervised learning can often improve classification performance by utilizing unlabeled data for learning. In this paper, we proposed a novel approach to semi-supervised learning as as an optimization of both the classification energy and cluster compactness energy in the unlabeled dataset. The resulting integer programming problem is relaxed by a semi-definite relaxation where efficient solution can be obtained. Furthermore, the spectral graph methods provide improved energy minimization via the incorporation of additional criteria. Results on UCI datasets show promising results.",

keywords = "Spectral Method, Label Data, Unlabeled Data, Data Mining Problem, Transductive Support Vector Machine",

author = "Chun-hung Li and Zhi-li Wu",

year = "2004",

month = apr,

day = "22",

doi = "10.1007/978-3-540-24775-3_4",

language = "English",

isbn = "9783540220640",

series = "Lecture Notes in Computer Science",

publisher = "Springer Berlin Heidelberg",

pages = "13--21",

editor = "Honghua Dai and Ramakrishnan Srikant and Chengqi Zhang",

booktitle = "Advances in Knowledge Discovery and Data Mining",

edition = "1st",

note = "8th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining, PAKDD 2004, PAKDD 2004 ; Conference date: 26-05-2004 Through 28-05-2004",

url = "https://link.springer.com/book/10.1007/b97861",

}

Li, C & Wu, Z 2004, Spectral Energy Minimization for Semi-supervised Learning. in H Dai, R Srikant & C Zhang (eds), Advances in Knowledge Discovery and Data Mining: 8th Pacific-Asia Conference, PAKDD 2004, Sydney, Australia, May 26-28, 2004, Proceedings. 1st edn, Lecture Notes in Computer Science, vol. 3056, Lecture Notes in Artificial Intelligence, PAKDD: Pacific-Asia Conference on Knowledge Discovery and Data Mining, Springer Berlin Heidelberg, pp. 13-21, 8th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining, PAKDD 2004, Sydney, Australia, 26/05/04. https://doi.org/10.1007/978-3-540-24775-3_4

Spectral Energy Minimization for Semi-supervised Learning. / Li, Chun-hung; Wu, Zhi-li.
Advances in Knowledge Discovery and Data Mining: 8th Pacific-Asia Conference, PAKDD 2004, Sydney, Australia, May 26-28, 2004, Proceedings. ed. / Honghua Dai; Ramakrishnan Srikant; Chengqi Zhang. 1st. ed. Springer Berlin Heidelberg, 2004. p. 13-21 (Lecture Notes in Computer Science; Vol. 3056), (Lecture Notes in Artificial Intelligence), (PAKDD: Pacific-Asia Conference on Knowledge Discovery and Data Mining).

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

TY - GEN

T1 - Spectral Energy Minimization for Semi-supervised Learning

AU - Li, Chun-hung

AU - Wu, Zhi-li

PY - 2004/4/22

Y1 - 2004/4/22

N2 - Data mining problems often involve a large amount of unlabeled data and there is often very limited known information on the dataset. In such scenario, semi-supervised learning can often improve classification performance by utilizing unlabeled data for learning. In this paper, we proposed a novel approach to semi-supervised learning as as an optimization of both the classification energy and cluster compactness energy in the unlabeled dataset. The resulting integer programming problem is relaxed by a semi-definite relaxation where efficient solution can be obtained. Furthermore, the spectral graph methods provide improved energy minimization via the incorporation of additional criteria. Results on UCI datasets show promising results.

AB - Data mining problems often involve a large amount of unlabeled data and there is often very limited known information on the dataset. In such scenario, semi-supervised learning can often improve classification performance by utilizing unlabeled data for learning. In this paper, we proposed a novel approach to semi-supervised learning as as an optimization of both the classification energy and cluster compactness energy in the unlabeled dataset. The resulting integer programming problem is relaxed by a semi-definite relaxation where efficient solution can be obtained. Furthermore, the spectral graph methods provide improved energy minimization via the incorporation of additional criteria. Results on UCI datasets show promising results.

KW - Spectral Method

KW - Label Data

KW - Unlabeled Data

KW - Data Mining Problem

KW - Transductive Support Vector Machine

U2 - 10.1007/978-3-540-24775-3_4

DO - 10.1007/978-3-540-24775-3_4

M3 - Conference proceeding

SN - 9783540220640

T3 - Lecture Notes in Computer Science

SP - 13

EP - 21

BT - Advances in Knowledge Discovery and Data Mining

A2 - Dai, Honghua

A2 - Srikant, Ramakrishnan

A2 - Zhang, Chengqi

PB - Springer Berlin Heidelberg

T2 - 8th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining, PAKDD 2004

Y2 - 26 May 2004 through 28 May 2004

ER -

Li C, Wu Z. Spectral Energy Minimization for Semi-supervised Learning. In Dai H, Srikant R, Zhang C, editors, Advances in Knowledge Discovery and Data Mining: 8th Pacific-Asia Conference, PAKDD 2004, Sydney, Australia, May 26-28, 2004, Proceedings. 1st ed. Springer Berlin Heidelberg. 2004. p. 13-21. (Lecture Notes in Computer Science). (Lecture Notes in Artificial Intelligence). (PAKDD: Pacific-Asia Conference on Knowledge Discovery and Data Mining). doi: 10.1007/978-3-540-24775-3_4

Spectral Energy Minimization for Semi-supervised Learning

Abstract

Publication series

Conference

User-Defined Keywords

Access to Document

Fingerprint

Cite this