Prediction Farmer Exchange Rate: Comparative Method of Analysis Holt-Winters Eksponential Smooting and Seasonal ARIMA Harizahayu (a*), Amin Harahap (b), Muhammad Fathoni (b), Hari Sumardi (b)
(a*)Politeknik Negeri Medan
Jl. Almamater NO.1 Kampus USU Medan
*harizahayu[at]polmed.ac.id
(b)Universitas Labuhan Batu
rectorat[at]ulb.ac.id
(b)Politeknik Unggula LP3M
JL. Iskanda Muda NO.3 ABCDF, Medan
fathoni[at]lp3m.ac.id
(b)Universitas Bengkulu
Abstract
This study aimed to predict seasonal time series data using the Holt-Winters exponential smoothing additive model and Seasonal autoregressive integrated moving average (ARIMA). The data used in this study is data on farmer term of trade at Nort Sumatera in 2016-2020, the source of the data obtained from thes Social website of the Central Statistics Agency. The results showed that the comparing of the Holt-Winters exponential smoothing method and SARIMA on farmer term of trade from 2016 to 2020 contained trend patterns and seasonal patterns by firrst determining the initial values and smoothing parameters minimize forecasting errors and get forecasthing from the best model. The best model to prec farmer term of trade is SARIMA (2,1,1) (0,1,1)_12 because The model fits the observed data well and shows no residual autocorrelation. The results of forecasting farmer term of trade at Nort Sumatera in 2016-2020 have increased continuously every month.
Keywords: Holt-Winters eksponential smoothing, SARIMA, Error, Best model, Forecasting
