APPLICATION OF FACTOR ANALYSIS IN THE STUDY OF PRODUCTIVE CHARACTERISTICS IN SHEEP

Abstract

Modern biometric measurements which are carried out in animal husbandry in the
process of scientific research, breeding work have rather high dimensionality. Simultaneous
analysis of a large number of interrelated quantities is quite a difficult task. At the same time,
it is worth considering that the initial measurements have different dimensionality, and can
also vary significantly in variance of values. In this regard, the task of analyzing the structure
of connections between the measured biometric parameters of animals is relevant. In order to
solve this problem, the technique of factor analysis is used. This technique allows you to reduce the dimension of the original measurement vector and move to new variables (factors),
the correlation between which is zero. The obtained factors make it possible to combine the
initial parameters according to the principle of the highest correlation. In the presented article, biometric data, which are obtained for an animal sample in the amount of 40 head, are examined using factor analysis. The interpretation of the factor analysis results is carried out in detail. The absolute values of the factor loadings matrix are examined. Based on this study, latent factors have been identified that combine several initial parameters with maximum correlation. Unique parameters that are not included in any of the selected factors are determined.
Furthermore, the analysis of the resulting scores is carried out. Based on this analysis, animals
with the maximum and minimum level of the selected factors are determined. During the
preparation of the article and performing calculations, computer mathematical tools such as
the MS Excel spreadsheet, Matlab and SPSS Statistics mathematical software packages were
used.