Time Series Analysis for Customs Revenue Prediction using Arima Model in Python

Abstract

The Directorate General of Customs and Excise (DJBC) serves as a revenue collector in the field of customs and excise. This revenue plays an essential role in supporting infrastructure development. Predictions are needed to plan a good State Revenue and Expenditure Budget (APBN). Predictions serve as a tool for revenue optimization and control. However, forecasting is problematic because unpredictable external factors also influence these receipts. A logical and accountable approach is needed to predict acceptance to overcome this problem. The prediction method used is Autoregressive Integrated Moving Average (ARIMA). According to the computations, the Root Mean Square Percentage Error (RMSPE) value is less than 10%, indicating that the ARIMA model estimation is excellent