Sea transportation is an inseparable and indispensable part of society in everyday lifefor Indonesian people especially during special moment such as Eid al-Fitr. This can be shown by the increasing of the number of sea transport passengers duringEid al-Fitr every year. The month shift during Eid al-Fitr shows the effect of calendar variation.The calendar variation method is a method that combines the dummy regression model with the ARIMA model. The purpose of this research is to obtain the best model by using time series analysis approach on ARIMA Box-Jenkins method and calendar variationARIMAX method to predict the number of domestic sea passenger at Tanjung Priok Port for 12 periods in the future. Based on the analysison the data of the number of domestic sea passenger at the Port of Tanjung Priok, it is concluded that the method of calendar variationARIMAX as the best method with ARIMA model (0, 0, [3]), V2,t,S1,t, S2,t, S3,t, S4,t, S5,t, S6,t, S7,t, S8,t, S9,t, S10,t, S11,t,t, V1,tt, V2,tt, S7,tt,S8,tt, S9,ttbecause it has the smallest MAPE value that is 14.3782% which indicates that the result of forecasting is good. Keywords : Sea Passengers, ARIMA Box-Jenkins, Calendar Variation, ARIMAX
Copyrights © 2018