Modelling Occupancy Rates in Croatian Hotel Industry

Abstract

The present study aims to analyse and model the net occupancy rates of bed-places in the Croatian hotel industry in the period from January 2005 to August 2014. For these purposes three models of generating time-series forecasts were used: the seasonal Naive model, the Holt-Winters exponential model and the seasonal autoregressive integrated moving average model. The empirical results show that the used time-series models performed well in terms of MAPE and that the Holt-Winters model outperformed the seasonal naive and the seasonal ARIMA model.