Multi-SVM classifier system with piecewise interpolation

Boyang Li, Qiangwei Wang, Takayuki Furuzuki

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Several researchers have shown that multiple classifier systems can result in effective solutions to difficult real-world classification tasks. However, most of these approaches are easily influenced by noise, and the training datasets for local classifiers get easily imbalanced. One of the main reasons for this is that it is hard to guarantee that the centers of the subsets are close to the separation hyperplane, so that it is difficult to evenly distribute the samples in the two sides of the hyperplane. In order to solve this problem, we redefine the description of classifier modeling problem as a task of piecewise approximation of the separation hyperplane. On the basis of this description, we propose a novel multiple support vector machine (SVM) classifier system. Its main contribution is a novel construction approach to the subtraining datasets. The proposed approach partitions the area close to the separation hyperplane into some subsets to construct the subtraining datasets. The subtraining datasets describe the subtasks for identifying segments of the separation hyperplane. Local SVMs are trained to solve the respective subtasks. Finally, the decisions of these local SVMs are appropriately combined on the basis of a probabilistic interpretation to obtain the final classification decision. The effectiveness of this approach is demonstrated through comparisons with some well-known approaches on both synthetic and real-world datasets.

Original languageEnglish
Pages (from-to)132-138
Number of pages7
JournalIEEJ Transactions on Electrical and Electronic Engineering
Volume8
Issue number2
DOIs
Publication statusPublished - 2013 Mar

Fingerprint

Support vector machines
Interpolation
Classifiers

Keywords

  • Mean filter
  • Multiple classifier system
  • Piecewise interpolation
  • Support vector machine

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Cite this

Multi-SVM classifier system with piecewise interpolation. / Li, Boyang; Wang, Qiangwei; Furuzuki, Takayuki.

In: IEEJ Transactions on Electrical and Electronic Engineering, Vol. 8, No. 2, 03.2013, p. 132-138.

Research output: Contribution to journalArticle

@article{744daa37cba148dda94fc49a72b65e4a,
title = "Multi-SVM classifier system with piecewise interpolation",
abstract = "Several researchers have shown that multiple classifier systems can result in effective solutions to difficult real-world classification tasks. However, most of these approaches are easily influenced by noise, and the training datasets for local classifiers get easily imbalanced. One of the main reasons for this is that it is hard to guarantee that the centers of the subsets are close to the separation hyperplane, so that it is difficult to evenly distribute the samples in the two sides of the hyperplane. In order to solve this problem, we redefine the description of classifier modeling problem as a task of piecewise approximation of the separation hyperplane. On the basis of this description, we propose a novel multiple support vector machine (SVM) classifier system. Its main contribution is a novel construction approach to the subtraining datasets. The proposed approach partitions the area close to the separation hyperplane into some subsets to construct the subtraining datasets. The subtraining datasets describe the subtasks for identifying segments of the separation hyperplane. Local SVMs are trained to solve the respective subtasks. Finally, the decisions of these local SVMs are appropriately combined on the basis of a probabilistic interpretation to obtain the final classification decision. The effectiveness of this approach is demonstrated through comparisons with some well-known approaches on both synthetic and real-world datasets.",
keywords = "Mean filter, Multiple classifier system, Piecewise interpolation, Support vector machine",
author = "Boyang Li and Qiangwei Wang and Takayuki Furuzuki",
year = "2013",
month = "3",
doi = "10.1002/tee.21832",
language = "English",
volume = "8",
pages = "132--138",
journal = "IEEJ Transactions on Electrical and Electronic Engineering",
issn = "1931-4973",
publisher = "John Wiley and Sons Inc.",
number = "2",

}

TY - JOUR

T1 - Multi-SVM classifier system with piecewise interpolation

AU - Li, Boyang

AU - Wang, Qiangwei

AU - Furuzuki, Takayuki

PY - 2013/3

Y1 - 2013/3

N2 - Several researchers have shown that multiple classifier systems can result in effective solutions to difficult real-world classification tasks. However, most of these approaches are easily influenced by noise, and the training datasets for local classifiers get easily imbalanced. One of the main reasons for this is that it is hard to guarantee that the centers of the subsets are close to the separation hyperplane, so that it is difficult to evenly distribute the samples in the two sides of the hyperplane. In order to solve this problem, we redefine the description of classifier modeling problem as a task of piecewise approximation of the separation hyperplane. On the basis of this description, we propose a novel multiple support vector machine (SVM) classifier system. Its main contribution is a novel construction approach to the subtraining datasets. The proposed approach partitions the area close to the separation hyperplane into some subsets to construct the subtraining datasets. The subtraining datasets describe the subtasks for identifying segments of the separation hyperplane. Local SVMs are trained to solve the respective subtasks. Finally, the decisions of these local SVMs are appropriately combined on the basis of a probabilistic interpretation to obtain the final classification decision. The effectiveness of this approach is demonstrated through comparisons with some well-known approaches on both synthetic and real-world datasets.

AB - Several researchers have shown that multiple classifier systems can result in effective solutions to difficult real-world classification tasks. However, most of these approaches are easily influenced by noise, and the training datasets for local classifiers get easily imbalanced. One of the main reasons for this is that it is hard to guarantee that the centers of the subsets are close to the separation hyperplane, so that it is difficult to evenly distribute the samples in the two sides of the hyperplane. In order to solve this problem, we redefine the description of classifier modeling problem as a task of piecewise approximation of the separation hyperplane. On the basis of this description, we propose a novel multiple support vector machine (SVM) classifier system. Its main contribution is a novel construction approach to the subtraining datasets. The proposed approach partitions the area close to the separation hyperplane into some subsets to construct the subtraining datasets. The subtraining datasets describe the subtasks for identifying segments of the separation hyperplane. Local SVMs are trained to solve the respective subtasks. Finally, the decisions of these local SVMs are appropriately combined on the basis of a probabilistic interpretation to obtain the final classification decision. The effectiveness of this approach is demonstrated through comparisons with some well-known approaches on both synthetic and real-world datasets.

KW - Mean filter

KW - Multiple classifier system

KW - Piecewise interpolation

KW - Support vector machine

UR - http://www.scopus.com/inward/record.url?scp=84873584757&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84873584757&partnerID=8YFLogxK

U2 - 10.1002/tee.21832

DO - 10.1002/tee.21832

M3 - Article

VL - 8

SP - 132

EP - 138

JO - IEEJ Transactions on Electrical and Electronic Engineering

JF - IEEJ Transactions on Electrical and Electronic Engineering

SN - 1931-4973

IS - 2

ER -