Optimasi Koefisien Single dan Double Exponential Smoothing Menggunakan Solver untuk Prediksi Produksi Gula Nasiona

Abstract

Sugar has an important in the national economy and food security; however, consumption levels continues to rise due to population growth and industrial expansion. Fluctuations in national sugar production—often falling short of demand—have created supply gaps and necessitated reliance on imports. To support strategic planning and balance supply and demand, this study forecasts national sugar production using historical data spanning 40 periods (1985–2024) from Statistics Indonesia (BPS) publications. The study’s main contribution is the use of Single Exponential Smoothing (SES) and Double Exponential Smoothing (DES) with MS Excel Solver to optimize the α and β coefficients, rather than relying on trial-and-error approaches or omitting detailed coefficient selection. Forecast results were evaluated using the Mean Absolute Percentage Error (MAPE) metric; both methods demonstrated high forecasting accuracy based on MAPE classification standards. A comparative analysis revealed nearly comparable performance, with SES yielding a slightly lower MAPE than DES.