Presentation 2008-03-13
A hierarchical multi-class classification method based on error-correcting output coding
Yoshimi OYAMA, Takashi TAKENOUCHI, Shin ISHII,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) There have been studies of multi-class classification by combining multiple binary classifiers within the framework of error-correcting output coding. In these methods, the original multi-class classification problem is decomposed to multiple binary classification problems by using the code matrix prepared in advance. In this report, we propose a new method of using parts of a code matrix "hierarchically". We apply this method to various classification problems including synthetic datasets and datasets from UCI repository and results show that the proposed method is superior or comparative to other multi-class classification methods such as the Hamming decoding and the multi-class SVM.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Multi-class classifier / error-correcting output coding / code matrix / hierarchical method
Paper # NC2007-169
Date of Issue

Conference Information
Committee NC
Conference Date 2008/3/5(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Neurocomputing (NC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A hierarchical multi-class classification method based on error-correcting output coding
Sub Title (in English)
Keyword(1) Multi-class classifier
Keyword(2) error-correcting output coding
Keyword(3) code matrix
Keyword(4) hierarchical method
1st Author's Name Yoshimi OYAMA
1st Author's Affiliation Faculty of Engineering, Kyoto University()
2nd Author's Name Takashi TAKENOUCHI
2nd Author's Affiliation Graduate School of Information Science, Nara Institute of Science and Technology
3rd Author's Name Shin ISHII
3rd Author's Affiliation Faculty of Engineering, Kyoto University
Date 2008-03-13
Paper # NC2007-169
Volume (vol) vol.107
Number (no) 542
Page pp.pp.-
#Pages 6
Date of Issue