<![CDATA[Rice is one of the staples that must be fulfilled to support human survival. As a result, if the price of rice is instable, it can cause a decrease in people's purchasing power. Therefore, a system is needed that can forecast rice prices to help maintain food security. This study uses the Holt-Winters Additive method because it can be used to predict time series data that has trend and seasonal patterns. The optimum parameter is found using the Golden Section optimization method that minimize the MAPE value. The data used is the average monthly data of rice prices at the level of large trade (wholesale) Indonesia. The results showed that the data contained elements of trend and seasonality additives and obtained the best model with α = 0.999702, β = 0.059114, γ = 0.145618. The results of measuring the forecasting ability of the formed model show that the forecast results are close to the actual data and are evidenced by the MAPE out sample value of 7.006% which is include MAPE criteria < 10% so that the forecasting ability is very high. The forecast results for 2023 show that rice prices have fluctuated but the changes are not too significant.]]>
Copyrights © 2025
