Prediksi Hasil Pertanian Tanaman Pangan Menggunakan Metode Double Exponential Smoothing

Abstract

Satar Mese district is one of the kecamatans that produces various staple foods for Kabupaten Manggarai. Three of these are lowland rice, dryland rice, and corn. High population growth results in the demand for various staple foods also increasing. The local government has no tools to predict the amount of staple food needs in the following year. In this regard, this study aims to forecast the production of 3 types of agricultural crops that are the main staple foods in the area. The data used is the data of the last ten years. The prediction method used is Double Exponential Smoothing (DES). MAPE (Mean Absolute Percentage Error) is used to calculate the error value. The results showed that the alpha value of 0.1 is more appropriate in predicting corn production, as seen from the smaller MAPE value of 18.85%. While for lowland rice and dryland rice products using an alpha value of 0.5. The system built is dynamic. Users can predict the amount of production in other fields, such as secondary crops, livestock, which is part of the duties of the Agricultural Extension Center (BPP) in providing information on technology in agriculture, production facilities, financing and markets for the sale of agricultural products.