Principal Component and Clustering Analysis of Functional Traits in Swiss Dairy Cattle
The objective of the research was to investigate the relationship among functional traits (body condition score (BCS), milk yield (MY), milking speed (MS), dry matter intake (DMI) and body weight (BW)). Data were from multiparous dairy cows (n = 55) of Chamau research farm of the Swiss Federal Institute of Technology, Switzerland. Principal component analysis with correlation matrix was used to find the relationship among BCS, MY, MS, DMI, BW, and other fixed effects including breed, year at calving, season, parity and year-season interaction. It was found that for all functional traits first 4 principal components explained more than 70% of the total variation. It was found that trading loss of accuracy using principal components scores instead of explanatory variables benefited reduction of dimension of explanatory variables and broke collinearity. Clustering analysis was performed based on different linkage methods and results showed physiological relationships among functional traits; since the data were from an experimental farm where each cow was fed by her MY performance and hence MY was associated with MS and DMI in the same cluster. BCS is correlated with BW and all these functional traits are related with mean lactation curve.
Principal Component and Clustering Analysis of Functional Traits in Swiss Dairy Cattle
The objective of the research was to investigate the relationship among functional traits (body condition score (BCS), milk yield (MY), milking speed (MS), dry matter intake (DMI) and body weight (BW)). Data were from multiparous dairy cows (n = 55) of Chamau research farm of the Swiss Federal Institute of Technology, Switzerland. Principal component analysis with correlation matrix was used to find the relationship among BCS, MY, MS, DMI, BW, and other fixed effects including breed, year at calving, season, parity and year-season interaction. It was found that for all functional traits first 4 principal components explained more than 70% of the total variation. It was found that trading loss of accuracy using principal components scores instead of explanatory variables benefited reduction of dimension of explanatory variables and broke collinearity. Clustering analysis was performed based on different linkage methods and results showed physiological relationships among functional traits; since the data were from an experimental farm where each cow was fed by her MY performance and hence MY was associated with MS and DMI in the same cluster. BCS is correlated with BW and all these functional traits are related with mean lactation curve.
