Binary classifier fusion based on the basic decomposition methods

Jaepil Ko, Hyeran Byun

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

For a complex multiclass problem, it is common to construct the multiclass classifier by combining the outputs of several binary ones. The two basic methods for this purpose are known as one-per-class (OPC) and pairwise coupling (PWC) and their general form is error correcting output code (ECOC). In this paper, we review basic decomposition methods and introduce a new sequential fusion method based on OPC and PWC according to their properties. In the experiments, we compare our proposed method with each basic method and ECOC method. The experimental results show that our proposed method can improve significantly the classification accuracy on the real dataset.

Original languageEnglish
Pages (from-to)146-155
Number of pages10
JournalLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume2709
Publication statusPublished - 2003 Dec 1

Fingerprint

Classifier Fusion
Decomposition Method
Classifiers
Fusion reactions
Binary
Decomposition
Multi-class
Pairwise
Output
Experiments
Fusion
Classifier
Experimental Results
Experiment

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

@article{8fc2351f14c640d7ab0d8f25136dba71,
title = "Binary classifier fusion based on the basic decomposition methods",
abstract = "For a complex multiclass problem, it is common to construct the multiclass classifier by combining the outputs of several binary ones. The two basic methods for this purpose are known as one-per-class (OPC) and pairwise coupling (PWC) and their general form is error correcting output code (ECOC). In this paper, we review basic decomposition methods and introduce a new sequential fusion method based on OPC and PWC according to their properties. In the experiments, we compare our proposed method with each basic method and ECOC method. The experimental results show that our proposed method can improve significantly the classification accuracy on the real dataset.",
author = "Jaepil Ko and Hyeran Byun",
year = "2003",
month = "12",
day = "1",
language = "English",
volume = "2709",
pages = "146--155",
journal = "Lecture Notes in Computer Science",
issn = "0302-9743",
publisher = "Springer Verlag",

}

TY - JOUR

T1 - Binary classifier fusion based on the basic decomposition methods

AU - Ko, Jaepil

AU - Byun, Hyeran

PY - 2003/12/1

Y1 - 2003/12/1

N2 - For a complex multiclass problem, it is common to construct the multiclass classifier by combining the outputs of several binary ones. The two basic methods for this purpose are known as one-per-class (OPC) and pairwise coupling (PWC) and their general form is error correcting output code (ECOC). In this paper, we review basic decomposition methods and introduce a new sequential fusion method based on OPC and PWC according to their properties. In the experiments, we compare our proposed method with each basic method and ECOC method. The experimental results show that our proposed method can improve significantly the classification accuracy on the real dataset.

AB - For a complex multiclass problem, it is common to construct the multiclass classifier by combining the outputs of several binary ones. The two basic methods for this purpose are known as one-per-class (OPC) and pairwise coupling (PWC) and their general form is error correcting output code (ECOC). In this paper, we review basic decomposition methods and introduce a new sequential fusion method based on OPC and PWC according to their properties. In the experiments, we compare our proposed method with each basic method and ECOC method. The experimental results show that our proposed method can improve significantly the classification accuracy on the real dataset.

UR - http://www.scopus.com/inward/record.url?scp=35248838888&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=35248838888&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:35248838888

VL - 2709

SP - 146

EP - 155

JO - Lecture Notes in Computer Science

JF - Lecture Notes in Computer Science

SN - 0302-9743

ER -