Accurate forecasting models in predicting the inbound tourism demand in Vietnam

Abstract

Tourism has been considered a complexly integrated and self-contained economic activity; however, it is one of the biggest industries in many countries. In order to make the tourism industry grow stably, it has been always an important issue to predict the tourism demand accurately. Though there have been many different studies in the methodologies, methods and models to forecast the tourism demand, there is no standard forecasting model that can be applied in different situations of the industry. In this study, two conventional models named ARIMA and Grey forecasting GM(1,1) were investigated. Their residuals were modified with Fourier series in order to improve the model accuracy level. In the empirical study of inbound tourism demand in Vietnam, the Fourier modified models called FARIMA(2,1,1) (1,0,2)12 and FGM(1,1) have very low values of mean absolute percentage error (MAPE) of 0.0055 and 0.0105, respectively. Both of them are excellent to forecast the inbound tourism demand in Vietnam but FARIMA(2,1,1)(1,0,2)12 is better and it is therefore suggested.