PENGUKURAN KINERJA METODE PERAMALAN TIPE EXPONENTIAL SMOOTHING DALAM PARAMETER TERBAIKNYA

Abstract

Abstrak Tipe exponential smoothing yang dikaji pada penelitian ini adalah simple exponential smoothing, double exponential smoothing, double exponential smoothing dengan damped trend dan triple exponential smoothing. Abstract The types of exponential smoothing studied in this study are simple exponential smoothing, double exponential smoothing, double exponential smoothing with damped trend and triple exponential smoothing. The purpose of this study is to find out the best exponential smoothing type performance in predicting the daily exchange rate of the rupiah against the USD. The performance of the exponential smoothing type forecasting method is very dependent on the selection of the smoothing parameter values contained in the mathematical equation. Therefore, the best parameter values for each type of exponential smoothing will be searched using the solver program. Getting the best parameter values is an optimization problem. In this study, the problem of optimizing the value of this parameter is to minimize forecasting errors based on MAPE with the constraint function is the forecast equation for each type of exponential smoothing. The stages of the study were divided into 2 segments, namely the training phase and the testing phase. The training phase produces the best smoothing parameter values. The best parameter values are then used as input for the testing phase. This testing phase is to determine the performance of each type of exponential smoothing. Exponential smoothing type performance is measured based on the magnitude of forecasting errors using MAPE, MAD and RMSE. The test results show that the performance of the double exponential smoothing method with the best damped trend compared to other exponential types with MAPE = 0.840, MAD = 117.28 and RMSE = 186.07.