Forecasting With ARIMA and SARIMA
Forecasting prices of basic commodities in Indonesia with Python and Jupyter Notebook
using Autoregressive Integrated Moving Average (ARIMA) and
Seasonal Autoregressive Integrated Moving Average (SARIMA).
Technology
Jupyter Notebook
Python
Description
This research predicts the price of chicken meat and chicken eggs using the ARIMA and SARIMA methods. stationary test using Rolling Mean and Rolling Standard Deviation visualization and Dickey-Duller Test, parameter optimization using the Grid Search method and diagnostic tests on the model using Ljung-Box test. Price forecasting is done on chicken meat and chicken eggs because the two commodities are interrelated with each other, seen from the results of the Pearson correlation test of 0.92 in the dataset and 0.87 in the forecasting results. The selection of the best model is based on the smallest Mean Squared Error (MSE), Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE). The best chicken meat price forecasting results using the ARIMA configuration method (3, 1, 2) with a MAPE value of 2.31%, while the best chicken egg forecasting results use the SARIMA configuration method ((2, 1, 1), (2, 0 , 2, 0), n] with MAPE values of 3.44%.