Multi-method ensemble selection of spectral bands related to leaf biochemistry

Research output: Contribution to journalArticle

54 Citations (Scopus)

Abstract

Multi-method ensembles are generally believed to return more reliable results than the application of one method alone. Here, we test if for the quantification of leaf traits an ensemble of regression models, consisting of Partial Least Squares (PLSR), Random Forest (RFR), and Support Vector Machine regression (SVMR) models, is able to improve the robustness of the spectral band selection process compared to the outcome of a single technique alone. The ensemble approach was tested using one artificial and five measured data sets of leaf level spectra and corresponding information on leaf chlorophyll, dry matter, and water content. PLSR models optimized for the goodness of fit, an established approach for band selection, were used to evaluate the performance of the ensemble. Although the fits of the models within the ensemble were poorer than the fits achieved with the reference approach, the ensemble was able to provide a band selection with higher consistency across all data sets. Due to the selection characteristics of the methods within the ensemble, the ensemble selection is moderately narrow and restrictive but in good agreement with known absorption features published in literature. We conclude that analyzing the range of agreement of different model types is an efficient way to select a robust set of spectral bands related to the foliar properties under investigation. This may help to deepen our understanding of the spectral response of biochemical and biophysical traits in foliage and canopies.

Original languageEnglish (US)
Pages (from-to)57-65
Number of pages9
JournalRemote Sensing of Environment
Volume164
DOIs
StatePublished - Jul 1 2015
Externally publishedYes

Fingerprint

Biochemistry
biochemistry
selection methods
leaves
least squares
Chlorophyll
dry matter content
methodology
Water content
foliage
Support vector machines
dry matter
chlorophyll
water content
method
spectral band
canopy
testing

Keywords

  • Hyperspectral
  • Imaging spectroscopy
  • Partial least squares regression
  • Random forest regression
  • Remote sensing
  • Support vector machine regression

ASJC Scopus subject areas

  • Soil Science
  • Geology
  • Computers in Earth Sciences

Cite this

Multi-method ensemble selection of spectral bands related to leaf biochemistry. / Feilhauer, Hannes; Asner, Gregory P.; Martin, Roberta E.

In: Remote Sensing of Environment, Vol. 164, 01.07.2015, p. 57-65.

Research output: Contribution to journalArticle

@article{423a2692697042009af271ff0427f5c6,
title = "Multi-method ensemble selection of spectral bands related to leaf biochemistry",
abstract = "Multi-method ensembles are generally believed to return more reliable results than the application of one method alone. Here, we test if for the quantification of leaf traits an ensemble of regression models, consisting of Partial Least Squares (PLSR), Random Forest (RFR), and Support Vector Machine regression (SVMR) models, is able to improve the robustness of the spectral band selection process compared to the outcome of a single technique alone. The ensemble approach was tested using one artificial and five measured data sets of leaf level spectra and corresponding information on leaf chlorophyll, dry matter, and water content. PLSR models optimized for the goodness of fit, an established approach for band selection, were used to evaluate the performance of the ensemble. Although the fits of the models within the ensemble were poorer than the fits achieved with the reference approach, the ensemble was able to provide a band selection with higher consistency across all data sets. Due to the selection characteristics of the methods within the ensemble, the ensemble selection is moderately narrow and restrictive but in good agreement with known absorption features published in literature. We conclude that analyzing the range of agreement of different model types is an efficient way to select a robust set of spectral bands related to the foliar properties under investigation. This may help to deepen our understanding of the spectral response of biochemical and biophysical traits in foliage and canopies.",
keywords = "Hyperspectral, Imaging spectroscopy, Partial least squares regression, Random forest regression, Remote sensing, Support vector machine regression",
author = "Hannes Feilhauer and Asner, {Gregory P.} and Martin, {Roberta E.}",
year = "2015",
month = "7",
day = "1",
doi = "10.1016/j.rse.2015.03.033",
language = "English (US)",
volume = "164",
pages = "57--65",
journal = "Remote Sensing of Environment",
issn = "0034-4257",
publisher = "Elsevier Inc.",

}

TY - JOUR

T1 - Multi-method ensemble selection of spectral bands related to leaf biochemistry

AU - Feilhauer, Hannes

AU - Asner, Gregory P.

AU - Martin, Roberta E.

PY - 2015/7/1

Y1 - 2015/7/1

N2 - Multi-method ensembles are generally believed to return more reliable results than the application of one method alone. Here, we test if for the quantification of leaf traits an ensemble of regression models, consisting of Partial Least Squares (PLSR), Random Forest (RFR), and Support Vector Machine regression (SVMR) models, is able to improve the robustness of the spectral band selection process compared to the outcome of a single technique alone. The ensemble approach was tested using one artificial and five measured data sets of leaf level spectra and corresponding information on leaf chlorophyll, dry matter, and water content. PLSR models optimized for the goodness of fit, an established approach for band selection, were used to evaluate the performance of the ensemble. Although the fits of the models within the ensemble were poorer than the fits achieved with the reference approach, the ensemble was able to provide a band selection with higher consistency across all data sets. Due to the selection characteristics of the methods within the ensemble, the ensemble selection is moderately narrow and restrictive but in good agreement with known absorption features published in literature. We conclude that analyzing the range of agreement of different model types is an efficient way to select a robust set of spectral bands related to the foliar properties under investigation. This may help to deepen our understanding of the spectral response of biochemical and biophysical traits in foliage and canopies.

AB - Multi-method ensembles are generally believed to return more reliable results than the application of one method alone. Here, we test if for the quantification of leaf traits an ensemble of regression models, consisting of Partial Least Squares (PLSR), Random Forest (RFR), and Support Vector Machine regression (SVMR) models, is able to improve the robustness of the spectral band selection process compared to the outcome of a single technique alone. The ensemble approach was tested using one artificial and five measured data sets of leaf level spectra and corresponding information on leaf chlorophyll, dry matter, and water content. PLSR models optimized for the goodness of fit, an established approach for band selection, were used to evaluate the performance of the ensemble. Although the fits of the models within the ensemble were poorer than the fits achieved with the reference approach, the ensemble was able to provide a band selection with higher consistency across all data sets. Due to the selection characteristics of the methods within the ensemble, the ensemble selection is moderately narrow and restrictive but in good agreement with known absorption features published in literature. We conclude that analyzing the range of agreement of different model types is an efficient way to select a robust set of spectral bands related to the foliar properties under investigation. This may help to deepen our understanding of the spectral response of biochemical and biophysical traits in foliage and canopies.

KW - Hyperspectral

KW - Imaging spectroscopy

KW - Partial least squares regression

KW - Random forest regression

KW - Remote sensing

KW - Support vector machine regression

UR - http://www.scopus.com/inward/record.url?scp=84928672637&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84928672637&partnerID=8YFLogxK

U2 - 10.1016/j.rse.2015.03.033

DO - 10.1016/j.rse.2015.03.033

M3 - Article

AN - SCOPUS:84928672637

VL - 164

SP - 57

EP - 65

JO - Remote Sensing of Environment

JF - Remote Sensing of Environment

SN - 0034-4257

ER -