HOLT WINTERS EXPONENTIAL SMOOTHING UNTUK MERAMALKAN PRODUK DOMESTIK BRUTO DI INDONESIA

A country's economic growth will be seen as having grown better or worse than in the past by measuring based on the increase in Gross Domestic Product (GDP). The pattern of Indonesian GDP from 2010 to 2022 shows that the data increases from year to year and there are seasonal fluctuations in the quarter. Holt Winters method is part of the Exponential Smoothing method used for forecasting if the data shows a trend and seasonality in the data pattern. The Holt Winters method has two models, namely additive and multiplicative. Holt Winters Additive is used if the data shows trends and seasonal patterns remain constant. Multiplicative Holt Winter is used if the data shows trends and seasonal patterns proportional to the average rate of the seasonal time series. The data used in this study are GDP Based on Current Prices (Nominal GDP) and GDP on the Basis of Constant Prices (Real GDP). Based on the evaluation of model performance using test data forecasting, the Holt Winters Multiplicative model of Nominal GDP with a MAPE value of 4,767535% is the best model because it has an accuracy value of <10%. While the Holt Winters Additive model of Real GDP with a MAPE value of 4,42387% is also the best model because it has an accuracy value of <10%.