<![CDATA[Forecasting is one effort that is often carried out by every institution with various business fields. Electricity consumption is a problem that requires a solution in the form of a forecasting model. The electricity consumption data is in the form of time series data which when used for forecasting can be in the form of univariate data or multivariate data. Forecasting using univariate data will be carried out in this study by utilizing the stacking method with the hope of producing a low RMSE value so that the accuracy of the research results becomes better. Besides that, the stacking method will also be applied to three different data intervals to find out the type of interval that is able to provide the smallest error value. Based on the research results, it was found that the application of stacking was able to reduce the RMSE value by 0.7%. In addition, the fact is also generated that the shortest data interval is better able to provide the smallest RMSE with a difference of about 7% with a longer interval.]]>
Copyrights © 2023
