Interpretation of Error Calculation Methods in the Context of Energy Forecasting
Forecasting is an inseparable part of any function of human activity and energetics is not an exception. After the privatization and deregulation of electric power systems, energy forecasting has become a popular subject which attracts the researchers in the relevant area. A myriad of studies are currently in existence in the literature about energy, load, or demand forecasting. The time interval of the forecasts may be in the horizon of very short-term, short- term, medium-term, or long-term forecasting; but almost in every one of them, analytical, artificial intelligence or hybrid techniques are employed. In all of the aforementioned terms above, one crucial thing is common and foremost in order to determine the performance of a forecasting technique within the specified time horizon, and that is the accuracy. There are various error calculation methods such as mean absolute percentage error (MAPE), mean squared error (MSE), mean absolute error (MAE), mean bias error (MBE) or different mathematical variations of the mentioned parameters. They are used as accuracy criteria to demonstrate the performance of a whole system or to compare with another analytical or artificial technique, although all error calculation methods have diversified meanings which correspond to particular phenomena. In this paper, the algebraic interpretation of error calculation methods in the context of energy forecasting is thoroughly presented in order to assist the prospective researchers in the field to use these methods properly.