Analisa Akurasi Penggunaan Metode Single Eksponential Smoothing Untuk Perkiraan Penjualan Minyak Solar (HSD)

Putri Octaria (Universitas Indo Global Mandiri)
Rangga Febri Kasih (Universitas Indo Global Mandiri)
Terttiavini (Universitas Indo Global Mandiri)

OKTAL : Jurnal Ilmu Komputer dan Sains
Vol 2 No 05 (2023): OKTAL : Jurnal Ilmu Komputer Dan Sains

This study analyzes the accuracy of the Single Exponential Smoothing (SES) method in predicting diesel oil (HSD) sales. The SES method is used in the oil and gas industry for inventory management, strategic decision making, financial planning, and market analysis. Historical diesel sales data is used to train and test the SES model in predicting future sales. The accuracy of the method is measured using metrics such as Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and Mean Absolute Percentage Error (MAPE). The forecasting method used is Single Exponential Smoothing, which uses the average value of past data to estimate future values. Three constant values of α (0.1, 0.6, and 0.9) were used, and the prediction results were evaluated using MAD, MSE, and MAPE. The results show that α = 0.1 gives the smallest MAPE, signifying higher accuracy in predicting HSD oil sales.