Novel feature selection method for genetic programming using metabolomic H-1 NMR data

R A Davis; A J Charlton; S Oehlschlager; J C Wilson

doi:10.1016/j.chemolab.2005.09.006

Novel feature selection method for genetic programming using metabolomic H-1 NMR data

R A Davis, A J Charlton, S Oehlschlager, J C Wilson

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

Abstract

A novel technique for multivariate data analysis using a two-stage genetic programming (GP) routine for feature selection is described. The method is compared with conventional genetic programming for the classification of genetically modified barley. Metabolic fingerprinting by H-1 NMR spectroscopy was used to analyse the differences between transgenic and null-segregant plants. We show that the method has a number of major advantages over standard genetic programming techniques. By selecting a minimal set of characteristic features in the data, the method provides models that are easier to interpret. Moreover the new method achieves better classification results and convergence is reached significantly faster. (c) 2005 Elsevier B.V All rights reserved.

Original language	English
Pages (from-to)	50-59
Number of pages	10
Journal	Chemometrics and Intelligent Laboratory Systems
Volume	81
Issue number	1
DOIs	https://doi.org/10.1016/j.chemolab.2005.09.006
Publication status	Published - 15 Mar 2006

Keywords

metabolomics
multivariate data analysis
genetic programming
feature selection
NMR
MULTIVARIATE-ANALYSIS
FUNCTIONAL GENOMICS
COMPLEX-MIXTURES
PLANT-EXTRACTS
SPECTRAL DATA
SPECTROSCOPY
CLASSIFICATION
CHEMOMETRICS
ALGORITHM
TOOL

Access to Document

10.1016/j.chemolab.2005.09.006

Cite this

@article{6b8e062d01c54ec98f06b1991ce95f19,

title = "Novel feature selection method for genetic programming using metabolomic H-1 NMR data",

abstract = "A novel technique for multivariate data analysis using a two-stage genetic programming (GP) routine for feature selection is described. The method is compared with conventional genetic programming for the classification of genetically modified barley. Metabolic fingerprinting by H-1 NMR spectroscopy was used to analyse the differences between transgenic and null-segregant plants. We show that the method has a number of major advantages over standard genetic programming techniques. By selecting a minimal set of characteristic features in the data, the method provides models that are easier to interpret. Moreover the new method achieves better classification results and convergence is reached significantly faster. (c) 2005 Elsevier B.V All rights reserved.",

keywords = "metabolomics, multivariate data analysis, genetic programming, feature selection, NMR, MULTIVARIATE-ANALYSIS, FUNCTIONAL GENOMICS, COMPLEX-MIXTURES, PLANT-EXTRACTS, SPECTRAL DATA, SPECTROSCOPY, CLASSIFICATION, CHEMOMETRICS, ALGORITHM, TOOL",

author = "Davis, {R A} and Charlton, {A J} and S Oehlschlager and Wilson, {J C}",

year = "2006",

month = mar,

day = "15",

doi = "10.1016/j.chemolab.2005.09.006",

language = "English",

volume = "81",

pages = "50--59",

journal = "Chemometrics and Intelligent Laboratory Systems",

issn = "0169-7439",

publisher = "Elsevier",

number = "1",

}

TY - JOUR

T1 - Novel feature selection method for genetic programming using metabolomic H-1 NMR data

AU - Davis, R A

AU - Charlton, A J

AU - Oehlschlager, S

AU - Wilson, J C

PY - 2006/3/15

Y1 - 2006/3/15

N2 - A novel technique for multivariate data analysis using a two-stage genetic programming (GP) routine for feature selection is described. The method is compared with conventional genetic programming for the classification of genetically modified barley. Metabolic fingerprinting by H-1 NMR spectroscopy was used to analyse the differences between transgenic and null-segregant plants. We show that the method has a number of major advantages over standard genetic programming techniques. By selecting a minimal set of characteristic features in the data, the method provides models that are easier to interpret. Moreover the new method achieves better classification results and convergence is reached significantly faster. (c) 2005 Elsevier B.V All rights reserved.

AB - A novel technique for multivariate data analysis using a two-stage genetic programming (GP) routine for feature selection is described. The method is compared with conventional genetic programming for the classification of genetically modified barley. Metabolic fingerprinting by H-1 NMR spectroscopy was used to analyse the differences between transgenic and null-segregant plants. We show that the method has a number of major advantages over standard genetic programming techniques. By selecting a minimal set of characteristic features in the data, the method provides models that are easier to interpret. Moreover the new method achieves better classification results and convergence is reached significantly faster. (c) 2005 Elsevier B.V All rights reserved.

KW - metabolomics

KW - multivariate data analysis

KW - genetic programming

KW - feature selection

KW - NMR

KW - MULTIVARIATE-ANALYSIS

KW - FUNCTIONAL GENOMICS

KW - COMPLEX-MIXTURES

KW - PLANT-EXTRACTS

KW - SPECTRAL DATA

KW - SPECTROSCOPY

KW - CLASSIFICATION

KW - CHEMOMETRICS

KW - ALGORITHM

KW - TOOL

U2 - 10.1016/j.chemolab.2005.09.006

DO - 10.1016/j.chemolab.2005.09.006

M3 - Article

SN - 0169-7439

VL - 81

SP - 50

EP - 59

JO - Chemometrics and Intelligent Laboratory Systems

JF - Chemometrics and Intelligent Laboratory Systems

IS - 1

ER -