<![CDATA[The process of predicting an event in the future is called forecasting. A forecasting model that functions to predict time series data with a trend pattern is Double Exponential Smoothing (DES). This study aims to compare one-parameter Brown DES with two-parameter Holt DES using the golden section method. The data used is monthly data on closing share prices of PT. Telkom Indonesia (Persero) for the period January 2011 - December 2021. Golden Section is an optimization method for finding parameter values that minimize the MAPE (Mean Absolute Percentage Error) function. The results of calculating the optimum parameter values for DES Brown α=0.420766 with a MAPE value of 4.871787804% and for DES Holt α=0.506578 and β=0.458980 with a MAPE value of 4.7233301647%. According to the MAPE value, the models used are very accurate for forecasting. DES Holt was selected as the best model for forecasting based on the smallest MAPE value.]]>
Copyrights © 2024
