| Abstract: | Tourism demand forecasting has attracted the attention of
researchers in the last decade. However, most of research
focused on traditional quantitative forecasting techniques,
such as ARIMA, exponential smoothing, etc. Although
these traditional methods have achieved certain levels
of success in the tourism research, it would be useful
to study the performance of alternative models such as
machine learning methods. This is the topic considered
in this paper. The goal is to investigate how different
machine learning models can be applied in the tourism
prediction problem and to assess the performance of seven
well known machine learning methods. Furthermore,
we investigate the effect of including the time index as
an input variable. Specifically, we consider the tourism
demand time series for Hong Kong inbound travel.
|