Optimizing prediction of human assessments of dairy odors using input variable selection

Fangle Chang, Paul H. Heinemann

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Use of instruments instead of human panels to assess odors can make the collection and measurement process more efficient and reliable. Odor-emitting samples from dairy farms, including manure, feed, and bedding materials, were collected and assessed by an electronic nose and a human panel. Artificial neural networks based on the Levenberg-Marquardt Back-propagation algorithm were used to build prediction models to predict human response to odor pleasantness. Feature selection methods, including Forward Selection (FS), Gamma Test (GT), and Principal Component Analysis (PCA), were applied to reduce the dimensionality of the measurements, potentially eliminating noise. Out of the 28 variable candidates (eNose sensors), 10 variables were selected when PCA was applied, and 16 variables were selected when either FS or GT approaches were applied. The model developed using GT provided the lowest mean square error of 0.56 (2.5%) hedonic scale units for separate validation. The GT-based model was able to predict the human assessments within 10% of the target for 81% of the independent validation samples and within 5% of the target for 63% of the independent validation samples.

Original languageEnglish (US)
Pages (from-to)402-410
Number of pages9
JournalComputers and Electronics in Agriculture
Volume150
DOIs
StatePublished - Jul 2018

Fingerprint

Dairies
Odors
odor
dairies
odors
Principal component analysis
prediction
principal component analysis
Backpropagation algorithms
Manures
testing
Mean square error
Farms
electronic nose
Feature extraction
back propagation
selection methods
sampling
dairy farming
artificial neural network

All Science Journal Classification (ASJC) codes

  • Forestry
  • Agronomy and Crop Science
  • Computer Science Applications
  • Horticulture

Cite this

@article{7a3f66868a7e46e4b1d0c7dbd7043c95,
title = "Optimizing prediction of human assessments of dairy odors using input variable selection",
abstract = "Use of instruments instead of human panels to assess odors can make the collection and measurement process more efficient and reliable. Odor-emitting samples from dairy farms, including manure, feed, and bedding materials, were collected and assessed by an electronic nose and a human panel. Artificial neural networks based on the Levenberg-Marquardt Back-propagation algorithm were used to build prediction models to predict human response to odor pleasantness. Feature selection methods, including Forward Selection (FS), Gamma Test (GT), and Principal Component Analysis (PCA), were applied to reduce the dimensionality of the measurements, potentially eliminating noise. Out of the 28 variable candidates (eNose sensors), 10 variables were selected when PCA was applied, and 16 variables were selected when either FS or GT approaches were applied. The model developed using GT provided the lowest mean square error of 0.56 (2.5{\%}) hedonic scale units for separate validation. The GT-based model was able to predict the human assessments within 10{\%} of the target for 81{\%} of the independent validation samples and within 5{\%} of the target for 63{\%} of the independent validation samples.",
author = "Fangle Chang and Heinemann, {Paul H.}",
year = "2018",
month = "7",
doi = "10.1016/j.compag.2018.05.017",
language = "English (US)",
volume = "150",
pages = "402--410",
journal = "Computers and Electronics in Agriculture",
issn = "0168-1699",
publisher = "Elsevier",

}

Optimizing prediction of human assessments of dairy odors using input variable selection. / Chang, Fangle; Heinemann, Paul H.

In: Computers and Electronics in Agriculture, Vol. 150, 07.2018, p. 402-410.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Optimizing prediction of human assessments of dairy odors using input variable selection

AU - Chang, Fangle

AU - Heinemann, Paul H.

PY - 2018/7

Y1 - 2018/7

N2 - Use of instruments instead of human panels to assess odors can make the collection and measurement process more efficient and reliable. Odor-emitting samples from dairy farms, including manure, feed, and bedding materials, were collected and assessed by an electronic nose and a human panel. Artificial neural networks based on the Levenberg-Marquardt Back-propagation algorithm were used to build prediction models to predict human response to odor pleasantness. Feature selection methods, including Forward Selection (FS), Gamma Test (GT), and Principal Component Analysis (PCA), were applied to reduce the dimensionality of the measurements, potentially eliminating noise. Out of the 28 variable candidates (eNose sensors), 10 variables were selected when PCA was applied, and 16 variables were selected when either FS or GT approaches were applied. The model developed using GT provided the lowest mean square error of 0.56 (2.5%) hedonic scale units for separate validation. The GT-based model was able to predict the human assessments within 10% of the target for 81% of the independent validation samples and within 5% of the target for 63% of the independent validation samples.

AB - Use of instruments instead of human panels to assess odors can make the collection and measurement process more efficient and reliable. Odor-emitting samples from dairy farms, including manure, feed, and bedding materials, were collected and assessed by an electronic nose and a human panel. Artificial neural networks based on the Levenberg-Marquardt Back-propagation algorithm were used to build prediction models to predict human response to odor pleasantness. Feature selection methods, including Forward Selection (FS), Gamma Test (GT), and Principal Component Analysis (PCA), were applied to reduce the dimensionality of the measurements, potentially eliminating noise. Out of the 28 variable candidates (eNose sensors), 10 variables were selected when PCA was applied, and 16 variables were selected when either FS or GT approaches were applied. The model developed using GT provided the lowest mean square error of 0.56 (2.5%) hedonic scale units for separate validation. The GT-based model was able to predict the human assessments within 10% of the target for 81% of the independent validation samples and within 5% of the target for 63% of the independent validation samples.

UR - http://www.scopus.com/inward/record.url?scp=85047066054&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85047066054&partnerID=8YFLogxK

U2 - 10.1016/j.compag.2018.05.017

DO - 10.1016/j.compag.2018.05.017

M3 - Article

AN - SCOPUS:85047066054

VL - 150

SP - 402

EP - 410

JO - Computers and Electronics in Agriculture

JF - Computers and Electronics in Agriculture

SN - 0168-1699

ER -