Permodelan Time Series Produksi Tanaman Padi di Provinsi Sumatera Utara

Abstract

Rice plants are plants that produce rice which is a source of carbohydrates, fat and protein for the body. Rice plants are used as a staple food for various levels of society. Specifically, the objectives of this research include building a time series model. Identify and develop a suitable time series model to predict rice production based on historical scheduled production data. Model performance test The performance test of the developed Time Series model uses data validation to convey the accuracy and precision of predictions. Comparison with conventional methods: Comparing the performance of time series models with conventional forecasting methods used in agricultural practice. Time Series Moving Average Method or MA (Exponential Smoothing) Method and MAPE (Mean Absolute Percentage Error) Method. Based on this model, the results of forecasting rice production from 2003 to 2023 experience ups and downs in each period.