Krakatau Steel is one of the manufacturing companies in Indonesia that has been experiencing losses for 9 years with the number of US$112 million in 2018. This research discusses forecasting discourse that many companies face while setting a new optimal strategic plan and aims to compare which method is best to use in terms of predicting krakatau steel demand. Data of accumulative steel products were collected from Krakatau Steel annual report with 14 years of historical data from 2005-2019. Double Exponential Smoothing and Single Exponential Smoothing methods were used to predict the demand in 2019. The accuracy was measured by MAPE, MAD, MSD, and MSE where the single exponential smoothing method is found as the best and more accurate than double exponential smoothing. Further improvements have been done to further examine the coefficient value and MAPE value towards the forecast results, still, Single Exponential Smoothing is the more accurate method.
Comparison of Double Exponential and Single Exponential Smoothing Accuracy in Krakatau Steel Demand Forecasting Fitted Model
1 Views
1 Downloads