Predictive analytics of Taiwan inbound tourism from ASEAN 5

Predictive analytics of Taiwan inbound tourism from ASEAN 5

This study evaluates Taiwan’s ‘New Southbound Policy’ initiative based on eight methods’ performances when predicting the country’s inbound tourism arrivals from ASEAN 5. We examine out-ofsample forecast accuracy by using a one-year holdout set and consider three main types of a hierarchical time series forecasting approach: bottom-up, top-down, and optimal combination. We then employ an aggregation of forecasts from each ARIMA model, the Holt-Winters exponential smoothing method, a time series regression, the ARIMA model, and the unobserved components model for aggregated tourism arrivals of ASEAN 5. We measure the forecast accuracy of these forecasting models/methods with four criteria by a rolling window approach. Based on the forecasts’ performance, we conclude that the ‘New Southbound Policy’ does help boost Taiwan’s tourism market. Moreover, the multiplicative Holt-Winters method ranks at the top based on a ranking system, while the unobserved components model exhibits the second-best rank.