PREDIKSI PRODUKSI BERAS DI PROVINSI JAWA TIMUR MENGGUNAKAN MUSIMAN AUTOREGRESSIVE TERINTEGRASI MOVING AVERAGE DENGAN METODE EXOGENOUS INPUT (SARIMAX)

Abstract

Indonesia is an agricultural country, meaning that agriculture plasy an important  role in the national economy. This can be proven by data from the Badan Pusat Statistik in 2017 which states that agriculture is one of the sectors that most contributes to the national economy with a contribution of 13,14 percent. Java has contributed around 54% of the national rice production. Based on BPS data in 2018, one of the largest rice producers in
Indonesia is Jawa Timur. With the land area that continues to decrease due to several factors, it is certain that rice production will decrease from year to year. One of the first steps in dealing with rice production in Jawa Timur is forecasting the coming year. The method used in this research is the Seansonal Autoregressive Integrated Moving Average
With Exogenous Input (SARIMAX). In this method, the factors affecting the dependent variabel Z at time t are influenced not only by the function of variable Z in time but also by other independent variables at time t. The results showed that the best model for predicting
rice production with the input variable rice harvest area was SARIMAX (1,0,0)(0,0,1) withthe transfer function parameter b=0, s=0, r=. . From the forecasting results wiht the best model, it is known that the highest riceproduction is always in the first subraund and the lowest is in the third subraund.

Keywords: Forecasting, Rice Production, SARIMAX.