Türk Tarım ve Doğa Bilimleri Dergisi

Yazarlar: Serhat ARSLAN, Mehmet Kazım KARA

Konular:-

DOI:10.30910/turkjans.448372

Anahtar Kelimeler:Random regression,DFREML,Repeated measurements,Orthogonal polinomial,Auto-correlation.

Özet: In animal breeding, repeated measurements are important. In this study, we compared the methods and models which are used in the analysis of data sets which contain the repeated measurements and the estimation of parameters. Morover, as an alternative method, random regression procedure which used the approach of covariance functions and was formed by compatibility of Ali-Schaeffer curve function was investigated. A data set was generated by simulation from retrospective records of dairy cattle. Fitting of the tested models for test-day yields in time were ranked from the best to the worst were CF-RRM (covariance function random regression model), DRRM (direct RRM), RM (repeatability model), ARM (auto-regresive model) and AM (animal model) respectively. It was determinded that the best models which explain the auto-correlation structure among the experimental errors were CF-RRM and ARM. Predicted parameters were compared for variances and the most sensitive estimation of parameters were obtained by CF-RRM. It was followed by DRRM.

---