Forecasting inbound tourism demand to China using time series models and belief functions

Abstract

Modeling uncertainty is a key issue in forecasting. In the tourism area, forecasts are used by governments, airline companies and operators to design tourism policies and they should include a quantification of uncertainties. This paper proposed a new approach to forecast the tourism demand, which is time series models combined with belief functions. We used this method to predict the demand for China international tourism, with an explicit representation of forecast uncertainty. The monthly data of international tourist arrival cover the period from January 1991 to June 2013. The result show that time seriesmodels combined with belief functions is a computationally simple and effective method.