Ensemble neural networks with novel gene-subsets for multiclass cancer classification

Jin Hyuk Hong, Sung Bae Cho

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Multiclass gene selection and classification of cancer are rapidly gaining attention in recent years, while conventional rank-based gene selection methods depend on predefined ideal marker genes that basically devised for binary classification. In this paper, we propose a novel gene selection method based on a gene's local class discriminability, which does not require any ideal marker genes for multiclass classification. An ensemble classifier with multiple NNs is trained with the gene subsets. The Global Cancer Map (GCM) cancer dataset is used to verify the proposed method for comparisons with the conventional approaches.

Original languageEnglish
Title of host publicationNeural Information Processing - 14th International Conference, ICONIP 2007, Revised Selected Papers
Pages856-865
Number of pages10
EditionPART 2
DOIs
Publication statusPublished - 2008 Oct 23
Event14th International Conference on Neural Information Processing, ICONIP 2007 - Kitakyushu, Japan
Duration: 2007 Nov 132007 Nov 16

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 2
Volume4985 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other14th International Conference on Neural Information Processing, ICONIP 2007
CountryJapan
CityKitakyushu
Period07/11/1307/11/16

Fingerprint

Neural Network Ensemble
Cancer Classification
Multi-class Classification
Gene Selection
Genes
Gene
Neural networks
Cancer
Subset
Ensemble Classifier
Binary Classification
Multi-class
Verify
Classifiers

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Hong, J. H., & Cho, S. B. (2008). Ensemble neural networks with novel gene-subsets for multiclass cancer classification. In Neural Information Processing - 14th International Conference, ICONIP 2007, Revised Selected Papers (PART 2 ed., pp. 856-865). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4985 LNCS, No. PART 2). https://doi.org/10.1007/978-3-540-69162-4_89
Hong, Jin Hyuk ; Cho, Sung Bae. / Ensemble neural networks with novel gene-subsets for multiclass cancer classification. Neural Information Processing - 14th International Conference, ICONIP 2007, Revised Selected Papers. PART 2. ed. 2008. pp. 856-865 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2).
@inproceedings{8f1d7ab8c4a94c9e9f918ae8ca6dd449,
title = "Ensemble neural networks with novel gene-subsets for multiclass cancer classification",
abstract = "Multiclass gene selection and classification of cancer are rapidly gaining attention in recent years, while conventional rank-based gene selection methods depend on predefined ideal marker genes that basically devised for binary classification. In this paper, we propose a novel gene selection method based on a gene's local class discriminability, which does not require any ideal marker genes for multiclass classification. An ensemble classifier with multiple NNs is trained with the gene subsets. The Global Cancer Map (GCM) cancer dataset is used to verify the proposed method for comparisons with the conventional approaches.",
author = "Hong, {Jin Hyuk} and Cho, {Sung Bae}",
year = "2008",
month = "10",
day = "23",
doi = "10.1007/978-3-540-69162-4_89",
language = "English",
isbn = "3540691596",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
number = "PART 2",
pages = "856--865",
booktitle = "Neural Information Processing - 14th International Conference, ICONIP 2007, Revised Selected Papers",
edition = "PART 2",

}

Hong, JH & Cho, SB 2008, Ensemble neural networks with novel gene-subsets for multiclass cancer classification. in Neural Information Processing - 14th International Conference, ICONIP 2007, Revised Selected Papers. PART 2 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 2, vol. 4985 LNCS, pp. 856-865, 14th International Conference on Neural Information Processing, ICONIP 2007, Kitakyushu, Japan, 07/11/13. https://doi.org/10.1007/978-3-540-69162-4_89

Ensemble neural networks with novel gene-subsets for multiclass cancer classification. / Hong, Jin Hyuk; Cho, Sung Bae.

Neural Information Processing - 14th International Conference, ICONIP 2007, Revised Selected Papers. PART 2. ed. 2008. p. 856-865 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4985 LNCS, No. PART 2).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Ensemble neural networks with novel gene-subsets for multiclass cancer classification

AU - Hong, Jin Hyuk

AU - Cho, Sung Bae

PY - 2008/10/23

Y1 - 2008/10/23

N2 - Multiclass gene selection and classification of cancer are rapidly gaining attention in recent years, while conventional rank-based gene selection methods depend on predefined ideal marker genes that basically devised for binary classification. In this paper, we propose a novel gene selection method based on a gene's local class discriminability, which does not require any ideal marker genes for multiclass classification. An ensemble classifier with multiple NNs is trained with the gene subsets. The Global Cancer Map (GCM) cancer dataset is used to verify the proposed method for comparisons with the conventional approaches.

AB - Multiclass gene selection and classification of cancer are rapidly gaining attention in recent years, while conventional rank-based gene selection methods depend on predefined ideal marker genes that basically devised for binary classification. In this paper, we propose a novel gene selection method based on a gene's local class discriminability, which does not require any ideal marker genes for multiclass classification. An ensemble classifier with multiple NNs is trained with the gene subsets. The Global Cancer Map (GCM) cancer dataset is used to verify the proposed method for comparisons with the conventional approaches.

UR - http://www.scopus.com/inward/record.url?scp=54049088217&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=54049088217&partnerID=8YFLogxK

U2 - 10.1007/978-3-540-69162-4_89

DO - 10.1007/978-3-540-69162-4_89

M3 - Conference contribution

AN - SCOPUS:54049088217

SN - 3540691596

SN - 9783540691594

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 856

EP - 865

BT - Neural Information Processing - 14th International Conference, ICONIP 2007, Revised Selected Papers

ER -

Hong JH, Cho SB. Ensemble neural networks with novel gene-subsets for multiclass cancer classification. In Neural Information Processing - 14th International Conference, ICONIP 2007, Revised Selected Papers. PART 2 ed. 2008. p. 856-865. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2). https://doi.org/10.1007/978-3-540-69162-4_89