Forecasting Tourist Arrivals to Sangiran Using Fuzzy with Calendar Variations

Abstract

Fuzzy method has been widely used in time series forecasting. However, the current fuzzy time models have not accommodated the holiday effects so that the forecasting error becomes large at certain moments. Regarding the problem, this study proposes two algorithms, extended of Chen's and seasonal fuzzy time series method (FTS), to consider the holiday effect in forecasting the monthly tourist arrivals to ancient human Sangiran Museum. Both algorithms consider the relationship between Eid holidays as the effect of calendar variations. The forecasting results obtained from the two proposed algorithms are then compared with those obtained from the Chen's and the seasonal FTS. Based on the experimental results, the proposed method can reduce mean absolute error (MAE), root mean square error (RMSE), and mean absolute percentage error (MAPE) obtained from Chen's method up to 61%, 61%, and 58%, respectively. Moreover, compared to that obtained from the seasonal FTS, the proposed method can reduce the MAE, RMSE, and MAPE values up to 35%, 36%, and 29%, respectively. The method proposed in this paper can be implemented to other time series with seasonal pattern and calendar variation effects.