Selecting appropriate nonlinear growth models using Bootstrap technique

Abstract

Growth is an important phase in the life of animals which influences the different forms of production such as milk, meat etc. The relationship between body weight and age is important particularly in meat producing animals. Many works have already been done for fitting of non linear growth models and choosing best model to describe growth pattern. In the present study, attention is given to the study of statistical properties of goodness of fit criteria for selecting best model to describe the growth pattern by using bootstrap technique. The distributions of the goodness of fit criteria R2 (determination coefficient), RMSE (root mean square error) and ARR (absolute reduction ratio) are found to be non-normal. Based on these statistical measures the best model is selected to describe growth pattern in given body weight data of goat. On comparison of 3 nonlinear growth models, viz. Logistic, Gompertz and Von bertalanffy model, the third one was found to be the best model.