Forecasting Hasil Produksi Perikanan Budidaya Laut Menggunakan ARIMA
The unstable production of yield production from marine aquaculture which is obtained every month, causing difficulties to determine the resources in the future. But this can be circumvented by making predictions or forecasting the yield production of marine aquaculture. Forecasting is done to plan operational of the UPT so it can satisfy requirements of the fishermen and consumers. One method that can be used is Autoregressive Integrated Moving Average (ARIMA) time series. This study aims to determine the forecasting model of marine aquaculture production with ARIMA, the data used in this study are data of marine aquaculture production (kg) by districts in East Java Province at 2015 - 2018. The accuracy test of the model is done by regard the MAPE value. The results is obtained by production model of marine aquaculture are Zt = - 0.0332 Zt-1 + 0.7119 Zt-2 + 0.5792 Zt-3 - 0.2585 Zt-4 + 0.3486 et-1 - 0.4424 et-2 - 0.9463 et-3 - 7.1287 with ARIMA (4, 0, 3), which means he model is feasible to be used to forecast the results of yield production of aquaculture fishery with 22.66% MAPE.