Seasonality and error autocorrelation in a Gompertz model adjustment for dairy cows weight growth

Abstract

In recent work (Kroll, 1990) on dairy cattle growth using its monthly weights, a Gompertz function, first considering the autocorrelated error structure, and second, the independent error structure, was fitted. The last explained 25% more of the total variance than the former. In this work, a Gompertz function was fitted with autocorrelated error structure adding seasonality to the model. This was suggested taking into account the time series autocorrelation function. In order to estimate the six parameters, an iterative procedure was used, employing a matrix =.[ | ], where is the usual matrix for fitting the Gompertz function, and is a 0s, 1s, -1s matrix necessary to take into account seasonality. With this model, the adjustment was better in terms of the coefficient determination.