Fonte: Cuadernos de Turismo; No. 48 (2021): Julio - Diciembre; 183-208
Palavras-chave:
Resumo: In the work, the time series 'tourists were visiting Andalusia' is modeled, a variable that has a strong seasonal component. The forecasting capacity of three different models is considered and analyzed, applying different modeling methodologies (Box-Jenkins, Holt-Winters and combined methods). The results obtained from the predictions are compared with the real values of the tourism series, assessing the good forecasting capacity of the three methodologies used. It is verified that the classic Holt-Winters procedure is the one that offers the best predictive results.