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 | |
| 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 |
Fingerprint
ASJC Scopus subject areas
- Computer Science(all)
- Theoretical Computer Science
Cite this
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
}
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
ER -