<![CDATA[Because the palm oil harvest cycle affects the seasonal pattern of crude palm oil (CPO) sales, operational planning strategically requires the ability to estimate sales volume. This study uses historical data from January 2020 to December 2025 to forecast monthly CPO sales volume using the multiplicative variation of the Holt-Winters Exponential Smoothing method. 80% of the data is used for training, and the remaining 20% is used for testing. Optimal parameters α=0.3742, β=0.0320, and γ=0.0000 are obtained through modeling. With an MAE value of 2,530.35 tons, an RMSE of 3,036.75 tons, and a MAPE of 12.24%, the model evaluation is classified as good. With a total forecast of 253,127.32 tons, the 12-month forecast results show the maximum peak in September 2026 at 26,743.34 tons and the lowest value in January 2026 at 15,091.27 tons.]]>
Copyrights © 2026
