PENGGUNAAN METODE PERAMALAN KOMBINASI TREND DETERMINISTIK DAN STOKASTIK PADA DATA JUMLAH PENUMPANG KERETA API (Studi Kasus: KA Argo Muria)

Abstract

The amount of the data of KA Argo Muria indicates the improve in every year during Ied mubarak day. Ied Mubarak day follows the Hijriyah calender, this is inditates that there is case effect of variation on the calender. The aims of this research is to predict the amount of the KA Argo Mulia passanger of destination of Semarang – Jakarta for 12 periodes in the future by using forecasting time series model of variation calender. The data used mounthly amount data KA Argo Mulia at PT KAI DAOP IV Semarang in the periode of January 2014 until Desember 2015. The result of the data analysis shows significant variable toward the model is D_(1,t) ,D_(2,t) ,D_(IdulFitri,t) ,D_(1,t) t,D_(10,t) t,D_(IdulFitri,t) t and the model of Autoregressive Integrated Moving Average (ARIMA) (1,0,0). Based on the result of forecasting out-sample data, is gained Mean Absolute Percentage Error (MAPE) is 1,8089 % which indicates that the result of forecasting is very good. Keywords: Deterministic trend, calendar variation, time series, stochastic, dummy regression.