Implementation of Forecasting Ginger Harvest using Seasonal Autoregressive Integrated Moving Average Method
Achmad Jauhari, Devie Rosa Anamisa and Fifin Ayu Mufarroha

University of Trunojoyo Madura

Abstract

Herbal plant farming is one of the holders of an essential role in the economy of Madura. Therefore, the Madurese government is very concerned about ginger farmers developing ginger production to meet market demand so that the Madurese economy increases. In addition, 2019 data from the Central Statistics Agency show that the herbal farming sector in Madura has been achieved by ginger farming with the most significant number of commodities compared to other herbal plants. However, in recent years, ginger yields have not been able to meet the very high market demand. Therefore, to meet consumer demand for the availability of ginger, this study uses forecasting analysis. The method used in this study is the Seasonal-Autoregressive Integrated Moving Average (SARIMA) hybrid method. This method is a pretty good method for modeling forecasting. The data used in this study are data on ginger production and harvested area from January 2015 to December 2019. And the results of this study are in the form of yield forecasting data for the following year. The test results with a data range of five produce small MAPE and RMSE, namely 43.94% and 14579.338. This shows that the SARIMA method has been able to predict future crop yields and can be used as a reference for the government in determining crop yields according to market demand.

Keywords: Forecasting, Ginger, SARIMA, Madura

Topic: Artificial Intelligence and Data Science