Learning when negative examples abound

Miroslav Kubat; Robert Holte; Stan Matwin

doi:10.1007/3-540-62858-4_79

Learning when negative examples abound

Miroslav Kubat, Robert Holte, Stan Matwin

Electrical and Computer Engineering

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

157 Scopus citations

Abstract

Existing concept learning systems can fail when the negative examples heavily outnumber the positive examples. The paper discusses one essential trouble brought about by imbalanced training sets and presents a learning algorithm addressing this issue. The experiments (with synthetic and real-world data) focus on 2-class problems with examples described with binary and continuous attributes.

Original language	English (US)
Title of host publication	Machine Learning
Subtitle of host publication	ECML-97 - 9th European Conference on Machine Learning, Proceedings
Editors	Maarten van Someren, Gerhard Widmer, Gerhard Widmer
Publisher	Springer Verlag
Pages	146-153
Number of pages	8
ISBN (Print)	3540628584, 9783540628583
DOIs	https://doi.org/10.1007/3-540-62858-4_79
State	Published - 1997
Event	9th European Conference on Machine Learning, ECML 1997 - Prague, Czech Republic Duration: Apr 23 1997 → Apr 25 1997

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	1224
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	9th European Conference on Machine Learning, ECML 1997
Country	Czech Republic
City	Prague
Period	4/23/97 → 4/25/97

ASJC Scopus subject areas

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/3-540-62858-4_79

Fingerprint Dive into the research topics of 'Learning when negative examples abound'. Together they form a unique fingerprint.

Cite this

Kubat, M., Holte, R., & Matwin, S. (1997). Learning when negative examples abound. In M. van Someren, G. Widmer, & G. Widmer (Eds.), Machine Learning: ECML-97 - 9th European Conference on Machine Learning, Proceedings (pp. 146-153). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1224). Springer Verlag. https://doi.org/10.1007/3-540-62858-4_79

Learning when negative examples abound. / Kubat, Miroslav; Holte, Robert; Matwin, Stan.

Machine Learning: ECML-97 - 9th European Conference on Machine Learning, Proceedings. ed. / Maarten van Someren; Gerhard Widmer; Gerhard Widmer. Springer Verlag, 1997. p. 146-153 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1224).

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

Kubat, M, Holte, R & Matwin, S 1997, Learning when negative examples abound. in M van Someren, G Widmer & G Widmer (eds), Machine Learning: ECML-97 - 9th European Conference on Machine Learning, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 1224, Springer Verlag, pp. 146-153, 9th European Conference on Machine Learning, ECML 1997, Prague, Czech Republic, 4/23/97. https://doi.org/10.1007/3-540-62858-4_79

Kubat M, Holte R, Matwin S. Learning when negative examples abound. In van Someren M, Widmer G, Widmer G, editors, Machine Learning: ECML-97 - 9th European Conference on Machine Learning, Proceedings. Springer Verlag. 1997. p. 146-153. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/3-540-62858-4_79

Kubat, Miroslav ; Holte, Robert ; Matwin, Stan. / Learning when negative examples abound. Machine Learning: ECML-97 - 9th European Conference on Machine Learning, Proceedings. editor / Maarten van Someren ; Gerhard Widmer ; Gerhard Widmer. Springer Verlag, 1997. pp. 146-153 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{71c6c139d3b0471f9ffcb358a4b03dee,

title = "Learning when negative examples abound",

abstract = "Existing concept learning systems can fail when the negative examples heavily outnumber the positive examples. The paper discusses one essential trouble brought about by imbalanced training sets and presents a learning algorithm addressing this issue. The experiments (with synthetic and real-world data) focus on 2-class problems with examples described with binary and continuous attributes.",

author = "Miroslav Kubat and Robert Holte and Stan Matwin",

note = "Publisher Copyright: {\textcopyright} Springer-Verlag Berlin Heidelberg 1997. Copyright: Copyright 2020 Elsevier B.V., All rights reserved.; 9th European Conference on Machine Learning, ECML 1997 ; Conference date: 23-04-1997 Through 25-04-1997",

year = "1997",

doi = "10.1007/3-540-62858-4_79",

language = "English (US)",

isbn = "3540628584",

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

publisher = "Springer Verlag",

pages = "146--153",

editor = "{van Someren}, Maarten and Gerhard Widmer and Gerhard Widmer",

booktitle = "Machine Learning",

}

TY - GEN

T1 - Learning when negative examples abound

AU - Kubat, Miroslav

AU - Holte, Robert

AU - Matwin, Stan

PY - 1997

Y1 - 1997

N2 - Existing concept learning systems can fail when the negative examples heavily outnumber the positive examples. The paper discusses one essential trouble brought about by imbalanced training sets and presents a learning algorithm addressing this issue. The experiments (with synthetic and real-world data) focus on 2-class problems with examples described with binary and continuous attributes.

AB - Existing concept learning systems can fail when the negative examples heavily outnumber the positive examples. The paper discusses one essential trouble brought about by imbalanced training sets and presents a learning algorithm addressing this issue. The experiments (with synthetic and real-world data) focus on 2-class problems with examples described with binary and continuous attributes.

UR - http://www.scopus.com/inward/record.url?scp=84947918649&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84947918649&partnerID=8YFLogxK

U2 - 10.1007/3-540-62858-4_79

DO - 10.1007/3-540-62858-4_79

M3 - Conference contribution

AN - SCOPUS:84947918649

SN - 3540628584

SN - 9783540628583

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

SP - 146

EP - 153

BT - Machine Learning

A2 - van Someren, Maarten

A2 - Widmer, Gerhard

PB - Springer Verlag

T2 - 9th European Conference on Machine Learning, ECML 1997

Y2 - 23 April 1997 through 25 April 1997

ER -