Show simple item record

dc.contributor.authorFrizzarin, Maria
dc.contributor.authorGormley, I. C.
dc.contributor.authorBerry, Donagh
dc.contributor.authorMurphy, T. B.
dc.contributor.authorCasa, A.
dc.contributor.authorLynch, A.
dc.contributor.authorMcParland, Sinead
dc.date.accessioned2021-03-16T17:10:34Z
dc.date.available2021-03-16T17:10:34Z
dc.date.issued2021
dc.identifier.citationFrizzarin, M. et al. Predicting cow milk quality traits from routinely available milk spectra using statistical machine learning methods.en_US
dc.identifier.urihttp://hdl.handle.net/11019/2390
dc.descriptionaccepted manuscripten_US
dc.description.abstractNumerous statistical machine learning methods suitable for application to highly correlated features, as exists for spectral data, could potentially improve prediction performance over the commonly used partial least squares approach. Milk samples from 622 individual cows with known detailed protein composition and technological trait data accompanied by mid-infrared spectra were available to assess the predictive ability of different regression and classification algorithms. The regression-based approaches were partial least squares regression (PLSR), ridge regression (RR), least absolute shrinkage and selection operator (LASSO), elastic net, principal component regression, projection pursuit regression, spike and slab regression, random forests, boosting decision trees, neural networks (NN) and a post-hoc approach of model averaging (MA). Several classification methods (i.e., partial least squares discriminant analysis (PLSDA), random forests, boosting decision trees, and support vector machines (SVM)) were also used after stratifying the traits of interest into categories. In the regression analyses, MA was the best prediction method for 6 of the 14 traits investigated (a60, alpha s1 CN, alpha s2 CN, kappa CN, alpha lactalbumin, and beta lactoglobulin B), while NN and RR were the best algorithms for 3 traits each (RCT, k20, and heat stability, and a30, beta CN, and beta lactoglobulin A, respectively), PLSR was best for pH and LASSO was best for CN micelle size. When traits were divided into two classes, SVM had the greatest accuracy for the majority of the traits investigated. While the well-established PLSR-based method performed competitively, the application of statistical machine learning methods for regression analyses reduced the root mean square error when compared to PLSR from between 0.18% (kappa CN) to 3.67% (heat stability). The use of modern statistical ML methods for trait prediction from MIRS may improve the prediction accuracy for some traits.en_US
dc.language.isoenen_US
dc.publisherElsevier for American Dairy Science Associationen_US
dc.relation.ispartofseriesJournal of Dairy Science;
dc.rightsAttribution-NonCommercial-ShareAlike 4.0 International*
dc.rights.urihttp://creativecommons.org/licenses/by-nc-sa/4.0/*
dc.subjectFourier transform mid-infrared spectroscopyen_US
dc.subjectstatistical machine learningen_US
dc.subjectmilk qualityen_US
dc.titlePredicting cow milk quality traits from routinely available milk spectra using statistical machine learning methods.en_US
dc.typeArticleen_US
dc.typeDataseten_US
dc.embargo.terms2022-06-01en_US
dc.contributor.sponsorScience Foundation Irelanden_US
dc.contributor.sponsorDepartment of Agriculture, Food and the Marineen_US
dc.contributor.sponsorGrantNumber18/SIRG/5562en_US
dc.contributor.sponsorGrantNumber16/RC/3835en_US


Files in this item

Thumbnail
Name:
AAM Frizzarin et al 2020- ...
Embargo:
2022-06-01
Size:
658.2Kb
Format:
PDF
Description:
main article
Thumbnail
Name:
MFrizzarin et al Supplemental ...
Embargo:
2022-06-01
Size:
532.9Kb
Format:
PDF
Description:
supplemental data

This item appears in the following Collection(s)

  • Livestock Systems [208]
    Teagasc LIvestock Systems Department includes Dairy, Cattle and Sheep research.

Show simple item record

Attribution-NonCommercial-ShareAlike 4.0 International
Except where otherwise noted, this item's license is described as Attribution-NonCommercial-ShareAlike 4.0 International