Optimasi Parameter Pemulusan Pada Metode Peramalan Double Exponential Smoothing Holt Menggunakan Golden Section

Abstract

Double Exponential Smoothing Holt (DES Holt) is a method that can be used when the data pattern shows a trend pattern. Determination of smoothing parameters usually uses trial and error, but this method still has inefficient results to get the best accuracy. One method that can be used to determine the smoothing parameters value is the golden section method. The application of the DES Holt and golden section methods will be carried out to predict the Exchange Rate of Farmers Subsector Livestock (ERFSL) of East Kalimantan Province. The purpose of this study was to obtain forecasting results and the level of accuracy of the ERFSL of East Kalimantan Province for the period January, February, and March 2020 using the DES Holt methods with the golden section smoothing parameter optimization method. The Forecasting results of DES Holt method have increased in the next three periods with an accuracy rate of 0.8856663%. The level of accuracy of forecasting results using the DES Holt methods has a MAPE value of less than 10%, which means the methods very good for predicting the ERFSL of East Kalimantan Province.