An error-counting network for pattern classification

Kar Ann Toh

doi:10.1016/j.neucom.2007.04.007

An error-counting network for pattern classification

Kar Ann Toh

Department of Electrical and Electronic Engineering

Research output: Contribution to journal › Article › peer-review

5 Citations (Scopus)

Abstract

This paper presents a novel quadratic error-counting network for pattern classification. Two computational issues namely, the network learning issue and the classification error-counting issue have been addressed. Essentially, a linear series functional approximation to network structure and a smooth quadratic error-counting cost function were proposed to resolve these two computational issues within a single framework. Our analysis shows that the quadratic error-counting objective can be related to the least-squares-error objective by adjusting the class-specific normalization factors. The binary classification network is subsequently extended to cater for multicategory problems. An extensive empirical evaluation validates the usefulness of proposed method.

Original language	English
Pages (from-to)	1680-1693
Number of pages	14
Journal	Neurocomputing
Volume	71
Issue number	7-9
DOIs	https://doi.org/10.1016/j.neucom.2007.04.007
Publication status	Published - 2008 Mar

Bibliographical note

Funding Information:
The author would like to thank Prof. Jaihie Kim and Prof. Sangyoun Lee for their kind support. Useful comments by anonymous reviewers are also acknowledged. This work was supported by the Korea Science and Engineering Foundation (KOSEF) through the Biometrics Engineering Research Center (BERC) at Yonsei University.

All Science Journal Classification (ASJC) codes

Computer Science Applications
Cognitive Neuroscience
Artificial Intelligence

Access to Document

10.1016/j.neucom.2007.04.007

Cite this

@article{6703c569d65c4a1a92e33597916616c8,

title = "An error-counting network for pattern classification",

abstract = "This paper presents a novel quadratic error-counting network for pattern classification. Two computational issues namely, the network learning issue and the classification error-counting issue have been addressed. Essentially, a linear series functional approximation to network structure and a smooth quadratic error-counting cost function were proposed to resolve these two computational issues within a single framework. Our analysis shows that the quadratic error-counting objective can be related to the least-squares-error objective by adjusting the class-specific normalization factors. The binary classification network is subsequently extended to cater for multicategory problems. An extensive empirical evaluation validates the usefulness of proposed method.",

author = "Toh, {Kar Ann}",

note = "Funding Information: The author would like to thank Prof. Jaihie Kim and Prof. Sangyoun Lee for their kind support. Useful comments by anonymous reviewers are also acknowledged. This work was supported by the Korea Science and Engineering Foundation (KOSEF) through the Biometrics Engineering Research Center (BERC) at Yonsei University. ",

year = "2008",

month = mar,

doi = "10.1016/j.neucom.2007.04.007",

language = "English",

volume = "71",

pages = "1680--1693",

journal = "Review of Economic Dynamics",

issn = "1094-2025",

publisher = "Academic Press Inc.",

number = "7-9",

}

TY - JOUR

T1 - An error-counting network for pattern classification

AU - Toh, Kar Ann

N1 - Funding Information: The author would like to thank Prof. Jaihie Kim and Prof. Sangyoun Lee for their kind support. Useful comments by anonymous reviewers are also acknowledged. This work was supported by the Korea Science and Engineering Foundation (KOSEF) through the Biometrics Engineering Research Center (BERC) at Yonsei University.

PY - 2008/3

Y1 - 2008/3

N2 - This paper presents a novel quadratic error-counting network for pattern classification. Two computational issues namely, the network learning issue and the classification error-counting issue have been addressed. Essentially, a linear series functional approximation to network structure and a smooth quadratic error-counting cost function were proposed to resolve these two computational issues within a single framework. Our analysis shows that the quadratic error-counting objective can be related to the least-squares-error objective by adjusting the class-specific normalization factors. The binary classification network is subsequently extended to cater for multicategory problems. An extensive empirical evaluation validates the usefulness of proposed method.

AB - This paper presents a novel quadratic error-counting network for pattern classification. Two computational issues namely, the network learning issue and the classification error-counting issue have been addressed. Essentially, a linear series functional approximation to network structure and a smooth quadratic error-counting cost function were proposed to resolve these two computational issues within a single framework. Our analysis shows that the quadratic error-counting objective can be related to the least-squares-error objective by adjusting the class-specific normalization factors. The binary classification network is subsequently extended to cater for multicategory problems. An extensive empirical evaluation validates the usefulness of proposed method.

UR - http://www.scopus.com/inward/record.url?scp=40649108086&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=40649108086&partnerID=8YFLogxK

U2 - 10.1016/j.neucom.2007.04.007

DO - 10.1016/j.neucom.2007.04.007

M3 - Article

AN - SCOPUS:40649108086

SN - 1094-2025

VL - 71

SP - 1680

EP - 1693

JO - Review of Economic Dynamics

JF - Review of Economic Dynamics

IS - 7-9

ER -

An error-counting network for pattern classification

Abstract

Bibliographical note

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this