<![CDATA[This study aims to forecast rice harvest yields in Deli Serdang Regency using the Singular Spectrum Analysis (SSA) method and evaluate the accuracy of the resulting model. The study employed a quantitative approach using secondary data on monthly rice harvest yields, measured in terms of dry milled grain (Gabah Kering Giling, GKG), from January 2020 to December 2024, comprising 60 observations. The data were divided into 48 in-sample observations for model development and 12 out-of-sample observations for testing. The analytical procedures included embedding, decomposition using Singular Value Decomposition (SVD), component grouping and reconstruction, and forecasting using the R-forecasting approach. The results indicate that the optimal window length among the candidates tested was (L = 17), with 17 eigentriples grouped into 13 groups. Evaluation on the test data yielded a Mean Absolute Percentage Error (MAPE) of 7.62%, indicating a relatively low overall forecasting error. The total forecast for 2025 was 517,310 metric tons, with the highest estimated yield in October at 52,901 metric tons and the lowest in March at 31,672 metric tons. These findings are expected to provide supporting information for production planning, harvest distribution, and food stock management in Deli Serdang Regency. However, the forecasts should be interpreted with consideration of variations in monthly errors and uncertainty regarding actual production levels.]]>
Copyrights © 2026
