Orman, Mehmet N.Çolak, Cemil2022-06-022022-06-022004http://hdl.handle.net/20.500.12575/80918Nonlinear Growth Models and The Study of Autocorrelation: An Application of Simmental X SAR Crossbred Cattle Genotypes This study was carried out to obtain growth curve models of the various body measurements of Simmental x Southern Anatolian Red (SAR) crossbred cattle Bi and FixBi genotypes. Also, it was aimed to inspect the autocorrelation of the residuals from the growth curve models and to eliminate the autocorrelation. The number of catties having Bi and FixBi genotypes were 8 and 52 heads respectively, totally 60 heads. The data of live weight, wither height, body length, chest circumference, chest depth, cannon bone circumference of these catties were recorded. Logistic, Gompertz, Richards and Monomolecular growth curve models were used to define the growth pattern. The invidual growth curve models of each genotypes were estimated for six variables explained above. The autocorrelation was detected in some of the residuals from the invidual growth curve models. The autocorrelation was eliminated and new models were estimated Median values of growth curve parameters were used to obtain general growth curve models. General growth curves were drawn for each genotypes. Goodness of fit of the growth curve models was determined by mean square error, determination coefficient and iteration number values. The values of determination coefficients of each models for all variables were estimated between 96-99,9 %. It was concluded that the best models of Bi genotype for live weight, wither height, body length, chest circumference, chest depth, cannon bone circumference were Logistic, Richards, Logistic, Gompertz, Gompertz and Logistic in sequence. The same arranging for FixBi genotype was Logistic, Logistic, Logistic, Richards, Richards and Logistic. Key Words: Growth model, nonlinear regression, autocorrelation, cattle, simmental, sar.trBiyoistatistikdoctoralThesis