Penerapan Metode Holt-Winters Exponential Smoothing Dalam Meramalkan Produksi Tanaman Pangan Nasional Tahun 2025

Abstract

Food crops play an essential role in meeting the basic needs of society, particularly rice, maize, and soybeans. This study aims to forecast national rice production by considering data patterns with trend and seasonal components. The method used is Holt-Winters Exponential Smoothing with a quantitative approach using secondary data. The analysis shows that the multiplicative Holt-Winters model is the best, with parameters α = 0.128, β = 0, and γ = 0.437. The model's Mean Absolute Percentage Error (MAPE) is 19.63%, indicating good forecasting accuracy in the range of 10-20%. The 2025 forecast shows an increase in rice production compared to previous years, suggesting improvements by stakeholders in maintaining future national food production stability.