Abstract
Support vector machines (SVMs), originally designed for binary classification, have been applied for multi-class classification, where an effective fusion scheme is required for combining outputs from them and producing a final result. In this work, we propose a novel method in which the SVMs are generated with the one-vs-rest (OVR) scheme and dynamically organized by the naive Bayes classifiers (NBs). This method might break the ties that frequently occur when working with multi-class classification systems with OVR SVMs. More specifically, we use the Pearson correlation measure to select informative genes and reduce the dimensionality of gene expression profiles when constructing the NBs. The proposed method has been validated on GCM cancer dataset consisting of 14 types of tumors with 16, 063 gene expression levels and produced higher accuracy than other methods.
| Original language | English |
|---|---|
| Title of host publication | Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings |
| Publisher | Springer Verlag |
| Pages | 155-164 |
| Number of pages | 10 |
| ISBN (Print) | 3540464840, 9783540464846 |
| Publication status | Published - 2006 Jan 1 |
| Event | 13th International Conference on Neural Information Processing, ICONIP 2006 - Hong Kong, China Duration: 2006 Oct 3 → 2006 Oct 6 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Volume | 4234 LNCS - III |
| ISSN (Print) | 0302-9743 |
| ISSN (Electronic) | 1611-3349 |
Other
| Other | 13th International Conference on Neural Information Processing, ICONIP 2006 |
|---|---|
| Country | China |
| City | Hong Kong |
| Period | 06/10/3 → 06/10/6 |
Fingerprint
All Science Journal Classification (ASJC) codes
- Theoretical Computer Science
- Computer Science(all)
Cite this
}
Multi-class cancer classification with OVR-support vector machines selected by Naïve bayes classifier. / Hong, Jin Hyuk; Cho, Sung Bae.
Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings. Springer Verlag, 2006. p. 155-164 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4234 LNCS - III).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
TY - GEN
T1 - Multi-class cancer classification with OVR-support vector machines selected by Naïve bayes classifier
AU - Hong, Jin Hyuk
AU - Cho, Sung Bae
PY - 2006/1/1
Y1 - 2006/1/1
N2 - Support vector machines (SVMs), originally designed for binary classification, have been applied for multi-class classification, where an effective fusion scheme is required for combining outputs from them and producing a final result. In this work, we propose a novel method in which the SVMs are generated with the one-vs-rest (OVR) scheme and dynamically organized by the naive Bayes classifiers (NBs). This method might break the ties that frequently occur when working with multi-class classification systems with OVR SVMs. More specifically, we use the Pearson correlation measure to select informative genes and reduce the dimensionality of gene expression profiles when constructing the NBs. The proposed method has been validated on GCM cancer dataset consisting of 14 types of tumors with 16, 063 gene expression levels and produced higher accuracy than other methods.
AB - Support vector machines (SVMs), originally designed for binary classification, have been applied for multi-class classification, where an effective fusion scheme is required for combining outputs from them and producing a final result. In this work, we propose a novel method in which the SVMs are generated with the one-vs-rest (OVR) scheme and dynamically organized by the naive Bayes classifiers (NBs). This method might break the ties that frequently occur when working with multi-class classification systems with OVR SVMs. More specifically, we use the Pearson correlation measure to select informative genes and reduce the dimensionality of gene expression profiles when constructing the NBs. The proposed method has been validated on GCM cancer dataset consisting of 14 types of tumors with 16, 063 gene expression levels and produced higher accuracy than other methods.
UR - http://www.scopus.com/inward/record.url?scp=33750717515&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=33750717515&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:33750717515
SN - 3540464840
SN - 9783540464846
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 155
EP - 164
BT - Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings
PB - Springer Verlag
ER -