Análise estatística multivariada de parâmetros de qualidade de leite cru refrigerado no estado de Minas Gerais

• Ano de defesa: 2008
• Autor(a) principal: Andrea Melo Garcia de Oliveira
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; UFMG
• Palavras-chave em Português: Qualidade de leite; Análise multivariada; Análise de componentes principais; Leite Análise
• Link de acesso: http://hdl.handle.net/1843/FRPO-7L2QZJ

Data analysis of refrigerated raw milk from 722 producers in five regions of the Minas Gerais State were used to evaluate, in a multivariate space, the associations between the variables to assess the milk quality. The following variables were considered: milk fat, protein, lactose, total solids, and solids nonfat contents, somatic cell count (SCC) and total bacterial count. The first analysis of principal components showed that from the seven principal components obtained from the correlation matrix, three had variance less than 0.7 (eigenvalue), which indicated the exclusion of three variables (those that have higher correlation with the principal components of less eigenvalue): total solids, and solids nonfat contents, and SCC. The exclusion of only two of these variables, total solids, and solids nonfat contents, resulted in higher correlation with levels of protein and fat. The third variable, SCC was not excluded because, internationally, it is one of the most important parameters to determine the quality of raw milk. Then, further analysis was performed and the associations between variables could be observed. With the implementation of each component a score was calculated for classification of producers. The cluster analysis was used to form groups according to similarity in quality of milk produced. It can be concluded that the approach of multivariate data analysis of raw milk is a good alternative as a tool to evaluate what are the most important variables to determine associations and to form group producer's according to milk quality.