Pronóstico de las condiciones del ganado vacuno en cuba a través de modelos ARIMA

The objective of this work was focused on modeling to predict animal conditions through the use of time series for Cuba. The analysis was based on the temperature and humidity index (ITH) calculated from the data of 64 meteorological stations of the country, grouped into three regions: east, center and west. Dickey - Fuller and KPSS tests were applied to identify the seasonal series. Holt Winters method of the family of time series of exponential smoothing was used. The best models were those corresponding to the eastern zone (1,1, 2) (0,0,1) 52 and (0,0,1) (0,1,1) 12. In addition, the analysis was carried out for a La Piedra weather station (Manicaragua, Villa Clara, Cuba).