Document Detail


New support vector-based design method for binary hierarchical classifiers for multi-class classification problems.
MedLine Citation:
PMID:  18187285     Owner:  NLM     Status:  MEDLINE    
Abstract/OtherAbstract:
We propose a new hierarchical design method, weighted support vector (WSV) k-means clustering, to design a binary hierarchical classification structure. This method automatically selects the classes to be separated at each node in the hierarchy, and allows visualization of clusters of high-dimensional support vector data; no prior hierarchical designs address this. At each node in the hierarchy, we use an SVRDM (support vector representation and discrimination machine) classifier, which offers generalization and good rejection of unseen false objects (rejection is not achieved with the standard SVMs). We give the basis and new insight into why a Gaussian kernel provides good rejection. Recognition and rejection test results on a real IR (infrared) database show that our proposed method outperforms the standard one-vs-rest methods and the use of standard SVM classifiers.
Authors:
Yu-Chiang Frank Wang; David Casasent
Related Documents :
12654315 - Multivariate statistics in analysis of data from the in vitro motility assay.
19831075 - Functional richness and ecosystem services: bird predation on arthropods in tropical ag...
17186815 - Classification using belief functions: relationship between case-based and model-based ...
3666675 - Classification of normal colorectal mucosa and adenocarcinoma by morphometry.
11796385 - Levofloxacin penetration into epithelial lining fluid as determined by population pharm...
18003205 - Parameter identifiability of cardiac ionic models using a novel cellml least squares op...
Publication Detail:
Type:  Journal Article     Date:  2007-12-08
Journal Detail:
Title:  Neural networks : the official journal of the International Neural Network Society     Volume:  21     ISSN:  0893-6080     ISO Abbreviation:  Neural Netw     Publication Date:    2008 Mar-Apr
Date Detail:
Created Date:  2008-03-17     Completed Date:  2008-07-10     Revised Date:  2008-08-15    
Medline Journal Info:
Nlm Unique ID:  8805018     Medline TA:  Neural Netw     Country:  United States    
Other Details:
Languages:  eng     Pagination:  502-10     Citation Subset:  IM    
Affiliation:
Department of Electrical and Computer Engineering, Carnegie Mellon University, Pittsburgh, PA 15213, USA. ycwang@cmu.edu
Export Citation:
APA/MLA Format     Download EndNote     Download BibTex
MeSH Terms
Descriptor/Qualifier:
Algorithms*
Artificial Intelligence*
Classification*
Databases, Factual
Discriminant Analysis
Fourier Analysis
Image Interpretation, Computer-Assisted
Information Storage and Retrieval
Least-Squares Analysis
Pattern Recognition, Automated / methods*
Comments/Corrections
Erratum In:
Neural Netw. 2008 May;21(4):698

From MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine


Previous Document:  Study of the relation between the incidence of uterine leiomyomas and the concentrations of PCB and ...
Next Document:  The 5-minute screening echocardiogram for athletes.