Multiple regression and machine learning based methods for carcass traits and saleable meat cuts prediction using non-invasive in vivo measurements in commercial lambs

Página do item simplificado
dc.contributor.authorAlves, Anderson Antonio Carvalho [UNESP]
dc.contributor.authorChaparro Pinzon, Andrés [UNESP]
dc.contributor.authorCosta, Rebeka Magalhães da [UNESP]
dc.contributor.authorSilva, Michelle Santos da
dc.contributor.authorVieira, Eloísa Helena Mendes
dc.contributor.authorMendonça, Ingrid Barbosa de
dc.contributor.authorViana, Vinícius de Sena Sales
dc.contributor.authorLôbo, Raimundo Nonato Braga
dc.contributor.institutionUniversidade Estadual Paulista (Unesp)
dc.contributor.institutionFederal University of Ceará (UFC)
dc.contributor.institutionEmpresa Brasileira de Pesquisa Agropecuária (EMBRAPA)
dc.date.accessioned2019-10-06T16:11:48Z
dc.date.available2019-10-06T16:11:48Z
dc.date.issued2019-02-01
dc.description.abstractThe aim of this study is to investigate the performance of multiple linear regression and machine learning methods to predict carcass traits and commercial meat cuts in lambs using non-invasive in vivo measurements. Bayesian networks were also investigated as an alternative method for feature selection. Phenotypes from 74 nondescript breed lambs were measured. A leave-one-out cross-validation strategy was performed and predictive ability statistics (R2, RMSE) were assessed. Moderate to high prediction accuracies were observed, with R2 values across models ranging, from 0.36 to 0.88 for carcass traits and 0.65 to 0.84 for meat cuts predictions. Results suggest that support vector machine algorithm is a potential alternative to the traditional multiple linear regression model. Further, the results of this study suggest that both stepwise and Bayesian network procedures could be useful as pre-selection tools of the input variables in non-parametric approaches and the best method for feature selection may be trait and model dependent.en
dc.description.affiliationDepartament of Animal Science State University of São Paulo (FCAV-UNESP) Rodovia de Acesso Prof. PauloDonato Castellane, S/N
dc.description.affiliationDepartment of Animal Science Federal University of Ceará (UFC), Av. Mister Hull, S/N
dc.description.affiliationEmbrapa Caprinos e Ovinos Estrada Sobral/Groaíras, km 04, Caixa postal 71
dc.description.affiliationUnespDepartament of Animal Science State University of São Paulo (FCAV-UNESP) Rodovia de Acesso Prof. PauloDonato Castellane, S/N
dc.format.extent49-56
dc.identifierhttp://dx.doi.org/10.1016/j.smallrumres.2018.12.008
dc.identifier.citationSmall Ruminant Research, v. 171, p. 49-56.
dc.identifier.doi10.1016/j.smallrumres.2018.12.008
dc.identifier.issn0921-4488
dc.identifier.scopus2-s2.0-85059171182
dc.identifier.urihttp://hdl.handle.net/11449/188551
dc.language.isoeng
dc.relation.ispartofSmall Ruminant Research
dc.rights.accessRightsAcesso aberto
dc.sourceScopus
dc.subjectArtificial neural network
dc.subjectBayesian network
dc.subjectCarcass weight
dc.subjectHot dressing percentage
dc.subjectSupport vector regression
dc.titleMultiple regression and machine learning based methods for carcass traits and saleable meat cuts prediction using non-invasive in vivo measurements in commercial lambsen
dc.typeArtigo
unesp.author.orcid0000-0001-8306-0487[1]
unesp.departmentZootecnia - FCAVpt

Arquivos

Coleções

Jaboticabal - FCAV - Faculdade de Ciências Agrárias e Veterinárias