PERBANDINGAN METODE ARIMA BOX-JENKIS DENGAN ARIMA ENSEMBLE PADA PERAMALAN NILAI EKSPOR PROVINSI BENGKULU

Abstract

Exports are the delivery of goods and services sold by residents of one country to residents of another country to obtain foreign currency from the purchasing country. TThe aim of the research is to find out the best model produced and compare the ARIMA Box-Jenskins method with the ARIMA Ensemble for forecasting monthly export value (millions of $) in Bengkulu province from January 2010 – December 2021. The data was analyzed using the ARIMA Box-Jenskins method with ARIMA Ensemble. The ARIMA method is a time series forecasting technique, using past values ​​of the dependent variable to make accurate short-term forecasts. ARIMA ensemble is a combination of forecast results from several ARIMA models which can be used to combine the output of different forecast results from ensemble members, namely ensemble averaging and ensemble stacking. From data processing, it was obtained that the best model used to predict the export value of Bengkulu province for the period January 2021 to December 2021 is the ARIMA(3,1,1) model because it has the smallest RMSE value, namely 6.242145 with the predicted value results for the next 12 periods.