Semi-supervised learning from only positive and unlabeled data using entropy

Xiaoling Wang; Zhen Xu; Chaofeng Sha; Martin Ester; Aoying Zhou

doi:10.1007/978-3-642-14246-8_64

Semi-supervised learning from only positive and unlabeled data using entropy

Xiaoling Wang^*, Zhen Xu, Chaofeng Sha, Martin Ester, Aoying Zhou

^*Corresponding author for this work

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

13 Scopus citations

Abstract

The problem of classification from positive and unlabeled examples attracts much attention currently. However, when the number of unlabeled negative examples is very small, the effectiveness of former work has been decreased. This paper propose an effective approach to address this problem, and we firstly use entropy to selects the likely positive and negative examples to build a complete training set; and then logistic regression classifier is applied on this new training set for classification. A series of experiments are conducted. The experimental results illustrate that the proposed approach outperforms previous work in the literature.

Original language	English
Title of host publication	Web-Age Information Management - 11th International Conference, WAIM 2010, Proceedings
Pages	668-679
Number of pages	12
DOIs	https://doi.org/10.1007/978-3-642-14246-8_64
State	Published - 2010
Event	11th International Conference on Web-Age Information Management, WAIM 2010 - Jiuzhaigou, China Duration: 15 Jul 2010 → 17 Jul 2010

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	6184 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	11th International Conference on Web-Age Information Management, WAIM 2010
Country/Territory	China
City	Jiuzhaigou
Period	15/07/10 → 17/07/10

Access to Document

10.1007/978-3-642-14246-8_64

Cite this

Wang, X., Xu, Z., Sha, C., Ester, M., & Zhou, A. (2010). Semi-supervised learning from only positive and unlabeled data using entropy. In Web-Age Information Management - 11th International Conference, WAIM 2010, Proceedings (pp. 668-679). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6184 LNCS). https://doi.org/10.1007/978-3-642-14246-8_64

Wang, Xiaoling ; Xu, Zhen ; Sha, Chaofeng et al. / Semi-supervised learning from only positive and unlabeled data using entropy. Web-Age Information Management - 11th International Conference, WAIM 2010, Proceedings. 2010. pp. 668-679 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{c2b5f38d1bbe47dabe476c50cca45450,

title = "Semi-supervised learning from only positive and unlabeled data using entropy",

abstract = "The problem of classification from positive and unlabeled examples attracts much attention currently. However, when the number of unlabeled negative examples is very small, the effectiveness of former work has been decreased. This paper propose an effective approach to address this problem, and we firstly use entropy to selects the likely positive and negative examples to build a complete training set; and then logistic regression classifier is applied on this new training set for classification. A series of experiments are conducted. The experimental results illustrate that the proposed approach outperforms previous work in the literature.",

author = "Xiaoling Wang and Zhen Xu and Chaofeng Sha and Martin Ester and Aoying Zhou",

year = "2010",

doi = "10.1007/978-3-642-14246-8\_64",

language = "英语",

isbn = "3642142451",

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

pages = "668--679",

booktitle = "Web-Age Information Management - 11th International Conference, WAIM 2010, Proceedings",

note = "11th International Conference on Web-Age Information Management, WAIM 2010 ; Conference date: 15-07-2010 Through 17-07-2010",

}

Wang, X, Xu, Z, Sha, C, Ester, M & Zhou, A 2010, Semi-supervised learning from only positive and unlabeled data using entropy. in Web-Age Information Management - 11th International Conference, WAIM 2010, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 6184 LNCS, pp. 668-679, 11th International Conference on Web-Age Information Management, WAIM 2010, Jiuzhaigou, China, 15/07/10. https://doi.org/10.1007/978-3-642-14246-8_64

Semi-supervised learning from only positive and unlabeled data using entropy. / Wang, Xiaoling; Xu, Zhen; Sha, Chaofeng et al.
Web-Age Information Management - 11th International Conference, WAIM 2010, Proceedings. 2010. p. 668-679 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6184 LNCS).

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

TY - GEN

T1 - Semi-supervised learning from only positive and unlabeled data using entropy

AU - Wang, Xiaoling

AU - Xu, Zhen

AU - Sha, Chaofeng

AU - Ester, Martin

AU - Zhou, Aoying

PY - 2010

Y1 - 2010

N2 - The problem of classification from positive and unlabeled examples attracts much attention currently. However, when the number of unlabeled negative examples is very small, the effectiveness of former work has been decreased. This paper propose an effective approach to address this problem, and we firstly use entropy to selects the likely positive and negative examples to build a complete training set; and then logistic regression classifier is applied on this new training set for classification. A series of experiments are conducted. The experimental results illustrate that the proposed approach outperforms previous work in the literature.

AB - The problem of classification from positive and unlabeled examples attracts much attention currently. However, when the number of unlabeled negative examples is very small, the effectiveness of former work has been decreased. This paper propose an effective approach to address this problem, and we firstly use entropy to selects the likely positive and negative examples to build a complete training set; and then logistic regression classifier is applied on this new training set for classification. A series of experiments are conducted. The experimental results illustrate that the proposed approach outperforms previous work in the literature.

UR - https://www.scopus.com/pages/publications/77955031833

U2 - 10.1007/978-3-642-14246-8_64

DO - 10.1007/978-3-642-14246-8_64

M3 - 会议稿件

AN - SCOPUS:77955031833

SN - 3642142451

SN - 9783642142451

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

SP - 668

EP - 679

BT - Web-Age Information Management - 11th International Conference, WAIM 2010, Proceedings

T2 - 11th International Conference on Web-Age Information Management, WAIM 2010

Y2 - 15 July 2010 through 17 July 2010

ER -

Wang X, Xu Z, Sha C, Ester M, Zhou A. Semi-supervised learning from only positive and unlabeled data using entropy. In Web-Age Information Management - 11th International Conference, WAIM 2010, Proceedings. 2010. p. 668-679. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-642-14246-8_64

Semi-supervised learning from only positive and unlabeled data using entropy

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this