جستجوی مقالات مرتبط با کلیدواژه "principal components analysis" در نشریات گروه "علوم دام"

The objective of the current study was to determine the live body weight regression equation using four traits including bodylength, heart circumference, height of withers and abdominal circumference in 150 camels of Yazdi breed. It was tried to describe the adequate details about an appropriate method for fixing or decreasing the problems of collinearity instability and correlation between independent variables related to live body weight of Yazdi camel ecotype. The results of this study showed that the problem of multicollinearity within the data was associated to body weight of Yazdi camel breed and related independent variables could be fixed by the principal components analysis. Among the independent variables, body height and abdominal circumference had the highest and lowest coefficient in estimating the body weight of Yazdi camel, respectively. It suggested that the principal components analysis be used when there is the problem of collinearity in multivariable linearly regression analysis because of more precise estimation than least squares method. Also this method can help to the breeders to select the best animals by predicting the precise value of some important traits and selecting the best independent variables for predicting the traits

The objective followed in the present study was to survey the relationship between 18 body trait measurements (live weight، height at wither، paunch girth، neck diameter، body length، girth around the body، width of fat tail at above، below and midpoint of fat tail، fat tail length lowers right and left sides، fat tail gap length، fat tail depth at the above، below، and midpoint، and girth around fat tail at the above، mid and down point) and the traits of fat tail weight and carcass (weight with and without fat tail) in Makooei Sheep and to Predict these traits، performance though multivariate Linear Regression Method Based on Principal Component Analysis. Sex showed significant effects on all the measured traits expect fat tail، depth at above، midpoint and down point، fat tail length towards right side plus gap length، width of fat tail at the above and below point of tail. Means and standard error for each trait were evaluated by sex. Multicollinearity was detected through a survey of the relationship among these traits، variance inflation factor and tolerance value. Principal component analysis was employed to resolve multicollinearity problem among independent variables and for a clearer explanation of the results. R2 range for different regression models varied between 0. 973 for carcass weight without tail in male sex and 0. 561 for carcass weight with fat tail in female sex of Makooei lambs.

The relationship between live body weight, body length, girth circumference, animal hight, upper, middle as well as lower width of fat-tail, fat-tail length, fat-tail gap length, fat-tail depth and fat-tail circumference along with fat-tail weight were determined using records of 731 Loribakhtiari sheep. Principal Component and Least Square Analyses were applied to solve the collinearity instability. Collinearity problems as portrayed by variance inflation factors above 5 or 10 were evident in some of independent variables. Results showed that the problem of collinearity in relation with fat-tail weight of 11 independent variables could be solved by using Principal Component Analysis method. Fat-tail gap length, fat-tail depth, and fat-tail circumference vs. girth circumference, and fat-tail length respectively represented the highest and the lowest coefficients regarding the estimation of fat-tail weight.