Logistic ensembles of Random Spherical Linear Oracles for microarray classification

Leif E. Peterson, Matthew A Coleman

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Random Spherical Linear Oracles (RSLO) for DNA microarray gene expression data are proposed for classifier fusion. RSLO employs random hyperplane splits of samples in the principal component score space based on the first three principal components (X, Y,Z) of the input feature set. Hyperplane splits are used to assign training(testing) samples to separate logistic regression mini-classifiers, which increases the diversity of voting results since errors are not shared across mini-classifiers. We recommend use of RSLO with 3-4 10-fold CV and re-partitioning samples randomly every ten iterations prior to each 10-fold CV. This equates to a total of 30-40 iterations.

Original languageEnglish (US)
Pages (from-to)382-397
Number of pages16
JournalInternational Journal of Data Mining and Bioinformatics
Volume3
Issue number4
DOIs
StatePublished - 2009
Externally publishedYes

Fingerprint

Politics
Microarrays
Oligonucleotide Array Sequence Analysis
Logistics
Classifiers
Logistic Models
logistics
Gene Expression
Gene expression
voting
DNA
Fusion reactions
regression
Testing

Keywords

  • Ensemble classifier fusion
  • Hyperplanes
  • Microarrays
  • PCs
  • Principal components
  • Principal directions
  • Random linear oracle
  • Random spherical linear oracles
  • RSLO

ASJC Scopus subject areas

  • Library and Information Sciences
  • Information Systems
  • Biochemistry, Genetics and Molecular Biology(all)

Cite this

Logistic ensembles of Random Spherical Linear Oracles for microarray classification. / Peterson, Leif E.; Coleman, Matthew A.

In: International Journal of Data Mining and Bioinformatics, Vol. 3, No. 4, 2009, p. 382-397.

Research output: Contribution to journalArticle

@article{5c786fbeb13c49669ac93ec0ff43a7a2,
title = "Logistic ensembles of Random Spherical Linear Oracles for microarray classification",
abstract = "Random Spherical Linear Oracles (RSLO) for DNA microarray gene expression data are proposed for classifier fusion. RSLO employs random hyperplane splits of samples in the principal component score space based on the first three principal components (X, Y,Z) of the input feature set. Hyperplane splits are used to assign training(testing) samples to separate logistic regression mini-classifiers, which increases the diversity of voting results since errors are not shared across mini-classifiers. We recommend use of RSLO with 3-4 10-fold CV and re-partitioning samples randomly every ten iterations prior to each 10-fold CV. This equates to a total of 30-40 iterations.",
keywords = "Ensemble classifier fusion, Hyperplanes, Microarrays, PCs, Principal components, Principal directions, Random linear oracle, Random spherical linear oracles, RSLO",
author = "Peterson, {Leif E.} and Coleman, {Matthew A}",
year = "2009",
doi = "10.1504/IJDMB.2009.029202",
language = "English (US)",
volume = "3",
pages = "382--397",
journal = "International Journal of Data Mining and Bioinformatics",
issn = "1748-5673",
publisher = "Inderscience Enterprises Ltd",
number = "4",

}

TY - JOUR

T1 - Logistic ensembles of Random Spherical Linear Oracles for microarray classification

AU - Peterson, Leif E.

AU - Coleman, Matthew A

PY - 2009

Y1 - 2009

N2 - Random Spherical Linear Oracles (RSLO) for DNA microarray gene expression data are proposed for classifier fusion. RSLO employs random hyperplane splits of samples in the principal component score space based on the first three principal components (X, Y,Z) of the input feature set. Hyperplane splits are used to assign training(testing) samples to separate logistic regression mini-classifiers, which increases the diversity of voting results since errors are not shared across mini-classifiers. We recommend use of RSLO with 3-4 10-fold CV and re-partitioning samples randomly every ten iterations prior to each 10-fold CV. This equates to a total of 30-40 iterations.

AB - Random Spherical Linear Oracles (RSLO) for DNA microarray gene expression data are proposed for classifier fusion. RSLO employs random hyperplane splits of samples in the principal component score space based on the first three principal components (X, Y,Z) of the input feature set. Hyperplane splits are used to assign training(testing) samples to separate logistic regression mini-classifiers, which increases the diversity of voting results since errors are not shared across mini-classifiers. We recommend use of RSLO with 3-4 10-fold CV and re-partitioning samples randomly every ten iterations prior to each 10-fold CV. This equates to a total of 30-40 iterations.

KW - Ensemble classifier fusion

KW - Hyperplanes

KW - Microarrays

KW - PCs

KW - Principal components

KW - Principal directions

KW - Random linear oracle

KW - Random spherical linear oracles

KW - RSLO

UR - http://www.scopus.com/inward/record.url?scp=70549087772&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70549087772&partnerID=8YFLogxK

U2 - 10.1504/IJDMB.2009.029202

DO - 10.1504/IJDMB.2009.029202

M3 - Article

C2 - 20052903

AN - SCOPUS:70549087772

VL - 3

SP - 382

EP - 397

JO - International Journal of Data Mining and Bioinformatics

JF - International Journal of Data Mining and Bioinformatics

SN - 1748-5673

IS - 4

ER -