Indonesia engages in bilateral international trade with Mozambique through the Indonesia-Mozambique Preferential Trade Agreement (IM-PTA). Data on import and export values obtained from the Central Statistics Agency (BPS) between January 2019 and July 2024 exhibits fluctuations. To forecast future import and export values, this research employs two data mining models: ARIMA and VAR. The goal is to identify the more accurate model for prediction purposes. The analysis results show that the ARIMA (1,1,0) model provides better predictions of export value with a Mean Absolute Percentage Error (MAPE) of 17.82 percent when compared to the VAR Model with a MAPE of 32.59 percent. However, both ARIMA and VAR models are unable to provide accurate predictions for import values ??because the MAPE results are more than 50 percent.
Copyrights © 2024