The purpose of this study is to forecast the arrival of foreign tourists in Indonesia by using Artificial Neural Network and Holt-Winters approach esutilising the historical data from January 2011 to December 2017. From the calculation process, we found that MAPE (Mean Absolute Percentage Error) value of Artificial Neural Network and HoltWinters approaches are and , respectively. So it can be concluded that the HoltWinters approach is better than the Artificial Neural Network approaches in forecasting the foreign tourists arrival in Indonesia. |