Forecasting the Indonesian Coffee Production and Consumption Using the Modified Golden Section Search to Estimate the Smoothing Parameters

The Double Exponential Smoothing (DES) and Triple (TES) are forecasting methods that require two three smoothing parameters, respectively. parameters often determined through a trial error process is not really efficient since many experiments need to be done. Therefore, in this study, parameter estimation algorithm conducted the form of modified Golden Section Search (GSS) obtain optimal from DES TES methods. Forecasting carried out on production, domestic consumption, export consumption data Indonesian coffee, which one leading agricultural sub-sector commodities. obtained Ministry Agriculture Republic Indonesia. by applying GSS used forecast production using method, while done with method. All MAPE values less than 20% indicates able perform good forecasting. results show coffee Indonesia cannot meet its demand until 2024 total exceeds production.