COMPARING NEURAL NETWORK AUTOREGRESSIVE METHOD FOR IMPORT DUTY REVENUE FORECASTING

Neural Network is one of the interesting methods in data analytics used to forecast time-series previous years. In initiative strategies, Directorate General Customs and Excise (DGCE) compares this method with Holt-Winters exponential smoothing get more accurate forecasting import duty revenue. This paper Network(NN) better for revenue using from Information System Automation (CEISA) billing system. As a result, NN gives result lower Mean Absolute Error (MAE), Percentage (MAPE), Root Square (RMSE). Therefore, neural networks should be monitor realization so DGCE can identify change economic indicators that affect earlier produce right policy respond it.