Peramalan Nilai Transaksi Uang Elektronik Di Indonesia Menggunakan Double Exponential Smoothing Brown Dengan Optimasi Nonlinier

Anna Putri Aritonang (Mulawarman University)
Meiliyani Siringoringo (Universitas Mulawarman)
Wiwit Pura Nurmayanti (Universitas Mulawarman)
Sri Wahyuningsih (Universitas Mulawarman)
Suyitno Suyitno (Universitas Mulawarman)

Eksponensial
Vol. 17 No. 1 (2026): Jurnal Eksponensial

The Double Exponential Smoothing (DES) Brown method is one of the forecasting methods used for data that exhibited a trend pattern, in which the smoothing process was performed twice. The determination of the optimal smoothing parameter in the DES Brown method is usually carried out through a trial-and-error process. Another way to obtain the optimal smoothing parameter value more quickly and accurately is by using nonlinear optimization. In this study, two optimization methods were used: the Golden Section and the Levenberg-Marquardt methods. The objectives of this research were to obtain the optimal smoothing parameter of the DES Brown method using the Golden Section and Levenberg-Marquardt optimizations, to forecast the value of electronic money transactions in Indonesia for the period of January to March 2025 using the DES Brown method with the optimal smoothing parameter, and to identify the best optimization method for determining the optimal smoothing parameter of DES Brown method were obtained based on the MAPE value. The results of the study showed that the optimal smoothing parameter of the DES Brown method using the Golden Section optimization was 0.4634178 and the Levenberg-Marquardt optimization was 0.3498674.