Discriminant analysis as a tool to classify farm hay in dairy farms.
Hay is one of the primary constituents of ruminant feed, and rapid classification systems of nutritional value are essential. A reliable approach to evaluating hay quality is a combination of visual combined inspection by NIRS analysis. The analysis was carried out on 1,639 samples of hay collected...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Public Library of Science (PLoS)
2023-01-01
|
| Series: | PLoS ONE |
| Online Access: | https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0294468&type=printable |
| _version_ | 1827587741504765952 |
|---|---|
| author | Aldo Dal Prà Riccardo Bozzi Silvia Parrini Alessandra Immovilli Roberto Davolio Fabrizio Ruozzi Maria Chiara Fabbri |
| author_facet | Aldo Dal Prà Riccardo Bozzi Silvia Parrini Alessandra Immovilli Roberto Davolio Fabrizio Ruozzi Maria Chiara Fabbri |
| author_sort | Aldo Dal Prà |
| collection | DOAJ |
| description | Hay is one of the primary constituents of ruminant feed, and rapid classification systems of nutritional value are essential. A reliable approach to evaluating hay quality is a combination of visual combined inspection by NIRS analysis. The analysis was carried out on 1,639 samples of hay collected from 2016 to 2021 in northern Italy. Discriminant analysis (DAPC) on five hay types (FOM, forage mixtures; APG, first alfalfa cutting with prevalence of graminaceous >50%; PRA, prevailing alfalfa >50%; PUA, purity alfalfa >95%; and PEM, permanent meadows) was performed by ex-ante visual inspection categorization and NIRS analysis. This study aimed to provide a complementary method to differentiate hay types and classify unknown samples. Two scenarios were used: i) all data were used for model training, and the discriminant functions were extracted based on all samples; ii) the assignment of each group was assessed without samples belonging to the training set group. DAPC model resulted in an overall assignment success rate of 66%; precisely, the success was 84, 79, 69, 37, and 27% for PUA, FOM, PRA, APG, and PEM, respectively. In the second scenario, three groups showed percentages of posterior assignment probability higher than 70% to only one group: PUA with PRA (~ 99%), PRA with PUA (~71%), and PEM with FOM (~75%). Discriminant analysis can be successfully used to differentiate hay types and could also be used to assess factors related to hay quality in addition to NIRS analysis. |
| first_indexed | 2024-03-09T00:22:23Z |
| format | Article |
| id | doaj.art-75bd5fd2194344068d3946e4703a9b80 |
| institution | Directory Open Access Journal |
| issn | 1932-6203 |
| language | English |
| last_indexed | 2024-03-09T00:22:23Z |
| publishDate | 2023-01-01 |
| publisher | Public Library of Science (PLoS) |
| record_format | Article |
| series | PLoS ONE |
| spelling | doaj.art-75bd5fd2194344068d3946e4703a9b802023-12-12T05:33:21ZengPublic Library of Science (PLoS)PLoS ONE1932-62032023-01-011811e029446810.1371/journal.pone.0294468Discriminant analysis as a tool to classify farm hay in dairy farms.Aldo Dal PràRiccardo BozziSilvia ParriniAlessandra ImmovilliRoberto DavolioFabrizio RuozziMaria Chiara FabbriHay is one of the primary constituents of ruminant feed, and rapid classification systems of nutritional value are essential. A reliable approach to evaluating hay quality is a combination of visual combined inspection by NIRS analysis. The analysis was carried out on 1,639 samples of hay collected from 2016 to 2021 in northern Italy. Discriminant analysis (DAPC) on five hay types (FOM, forage mixtures; APG, first alfalfa cutting with prevalence of graminaceous >50%; PRA, prevailing alfalfa >50%; PUA, purity alfalfa >95%; and PEM, permanent meadows) was performed by ex-ante visual inspection categorization and NIRS analysis. This study aimed to provide a complementary method to differentiate hay types and classify unknown samples. Two scenarios were used: i) all data were used for model training, and the discriminant functions were extracted based on all samples; ii) the assignment of each group was assessed without samples belonging to the training set group. DAPC model resulted in an overall assignment success rate of 66%; precisely, the success was 84, 79, 69, 37, and 27% for PUA, FOM, PRA, APG, and PEM, respectively. In the second scenario, three groups showed percentages of posterior assignment probability higher than 70% to only one group: PUA with PRA (~ 99%), PRA with PUA (~71%), and PEM with FOM (~75%). Discriminant analysis can be successfully used to differentiate hay types and could also be used to assess factors related to hay quality in addition to NIRS analysis.https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0294468&type=printable |
| spellingShingle | Aldo Dal Prà Riccardo Bozzi Silvia Parrini Alessandra Immovilli Roberto Davolio Fabrizio Ruozzi Maria Chiara Fabbri Discriminant analysis as a tool to classify farm hay in dairy farms. PLoS ONE |
| title | Discriminant analysis as a tool to classify farm hay in dairy farms. |
| title_full | Discriminant analysis as a tool to classify farm hay in dairy farms. |
| title_fullStr | Discriminant analysis as a tool to classify farm hay in dairy farms. |
| title_full_unstemmed | Discriminant analysis as a tool to classify farm hay in dairy farms. |
| title_short | Discriminant analysis as a tool to classify farm hay in dairy farms. |
| title_sort | discriminant analysis as a tool to classify farm hay in dairy farms |
| url | https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0294468&type=printable |
| work_keys_str_mv | AT aldodalpra discriminantanalysisasatooltoclassifyfarmhayindairyfarms AT riccardobozzi discriminantanalysisasatooltoclassifyfarmhayindairyfarms AT silviaparrini discriminantanalysisasatooltoclassifyfarmhayindairyfarms AT alessandraimmovilli discriminantanalysisasatooltoclassifyfarmhayindairyfarms AT robertodavolio discriminantanalysisasatooltoclassifyfarmhayindairyfarms AT fabrizioruozzi discriminantanalysisasatooltoclassifyfarmhayindairyfarms AT mariachiarafabbri discriminantanalysisasatooltoclassifyfarmhayindairyfarms |