Meta-classifiers and selective superiority

Ryan Benton; Miroslav Kubat; Rasaiah Loganantharaj

doi:10.1007/3-540-45049-1_53

Meta-classifiers and selective superiority

Ryan Benton, Miroslav Kubat, Rasaiah Loganantharaj

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

Abstract

Given that no one classification method is the best in all tasks, a variety of approaches have evolved to prevent poor performance due to mismatch of capabilities. One approach to overcome this problem is to determine when a method may be appropriate for a given problem. A second, more popular approach is to combine the capabilities of two or more classification methods. This paper provides some evidence that the combining of classifiers can yield more robust solutions.

Original language	English (US)
Title of host publication	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Publisher	Springer Verlag
Pages	434-442
Number of pages	9
Volume	1821
ISBN (Print)	3540676899, 9783540450498
DOIs	https://doi.org/10.1007/3-540-45049-1_53
State	Published - 2000
Externally published	Yes
Event	13th International Conference on Industrial and Engineering Applications of Artificial Intelligence and Expert Systems, IEA/AIE 2000 - New Orleans, United States Duration: Jun 19 2000 → Jun 22 2000

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	1821
ISSN (Print)	03029743
ISSN (Electronic)	16113349

Other

Other	13th International Conference on Industrial and Engineering Applications of Artificial Intelligence and Expert Systems, IEA/AIE 2000
Country	United States
City	New Orleans
Period	6/19/00 → 6/22/00

ASJC Scopus subject areas

Computer Science(all)
Theoretical Computer Science

Access to Document

10.1007/3-540-45049-1_53

Fingerprint Dive into the research topics of 'Meta-classifiers and selective superiority'. Together they form a unique fingerprint.

Cite this

Benton, R., Kubat, M., & Loganantharaj, R. (2000). Meta-classifiers and selective superiority. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1821, pp. 434-442). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1821). Springer Verlag. https://doi.org/10.1007/3-540-45049-1_53

Meta-classifiers and selective superiority. / Benton, Ryan; Kubat, Miroslav; Loganantharaj, Rasaiah.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 1821 Springer Verlag, 2000. p. 434-442 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1821).

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

Benton, R, Kubat, M & Loganantharaj, R 2000, Meta-classifiers and selective superiority. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 1821, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 1821, Springer Verlag, pp. 434-442, 13th International Conference on Industrial and Engineering Applications of Artificial Intelligence and Expert Systems, IEA/AIE 2000, New Orleans, United States, 6/19/00. https://doi.org/10.1007/3-540-45049-1_53

Benton R, Kubat M, Loganantharaj R. Meta-classifiers and selective superiority. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 1821. Springer Verlag. 2000. p. 434-442. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/3-540-45049-1_53

Benton, Ryan ; Kubat, Miroslav ; Loganantharaj, Rasaiah. / Meta-classifiers and selective superiority. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 1821 Springer Verlag, 2000. pp. 434-442 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{8d3b755932344fabb83a0842930cf218,

title = "Meta-classifiers and selective superiority",

abstract = "Given that no one classification method is the best in all tasks, a variety of approaches have evolved to prevent poor performance due to mismatch of capabilities. One approach to overcome this problem is to determine when a method may be appropriate for a given problem. A second, more popular approach is to combine the capabilities of two or more classification methods. This paper provides some evidence that the combining of classifiers can yield more robust solutions.",

author = "Ryan Benton and Miroslav Kubat and Rasaiah Loganantharaj",

year = "2000",

doi = "10.1007/3-540-45049-1_53",

language = "English (US)",

isbn = "3540676899",

volume = "1821",

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

publisher = "Springer Verlag",

pages = "434--442",

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

note = "13th International Conference on Industrial and Engineering Applications of Artificial Intelligence and Expert Systems, IEA/AIE 2000 ; Conference date: 19-06-2000 Through 22-06-2000",

}

TY - GEN

T1 - Meta-classifiers and selective superiority

AU - Benton, Ryan

AU - Kubat, Miroslav

AU - Loganantharaj, Rasaiah

PY - 2000

Y1 - 2000

N2 - Given that no one classification method is the best in all tasks, a variety of approaches have evolved to prevent poor performance due to mismatch of capabilities. One approach to overcome this problem is to determine when a method may be appropriate for a given problem. A second, more popular approach is to combine the capabilities of two or more classification methods. This paper provides some evidence that the combining of classifiers can yield more robust solutions.

AB - Given that no one classification method is the best in all tasks, a variety of approaches have evolved to prevent poor performance due to mismatch of capabilities. One approach to overcome this problem is to determine when a method may be appropriate for a given problem. A second, more popular approach is to combine the capabilities of two or more classification methods. This paper provides some evidence that the combining of classifiers can yield more robust solutions.

UR - http://www.scopus.com/inward/record.url?scp=84957869946&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84957869946&partnerID=8YFLogxK

U2 - 10.1007/3-540-45049-1_53

DO - 10.1007/3-540-45049-1_53

M3 - Conference contribution

AN - SCOPUS:84957869946

SN - 3540676899

SN - 9783540450498

VL - 1821

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

SP - 434

EP - 442

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

PB - Springer Verlag

T2 - 13th International Conference on Industrial and Engineering Applications of Artificial Intelligence and Expert Systems, IEA/AIE 2000

Y2 - 19 June 2000 through 22 June 2000

ER -