The paper examines the usefulness of the classical development trend method for forecasting motorization rate values. A method based on minimizing the ex ante forecast error was applied. By changing the number of the last L years included in the prediction process, the minimum value of the ex ante error is sought. For the optimal number of years, the forecast value is determined. The method was discussed using the example of predicting the values of selected motorization level indicators: the number of registered motor vehicles and passenger cars for selected cities in Poland: Gdańsk, Gliwice, Kraków, and Szczecin. Calculations were performed for data from the years 2000-2009, while data for 2010 were used to determine the ex post forecast error.The paper examines the usefulness of classical trend models for forecasting the values of selected motorization indicators. The following trend models were included in the prediction process: linear, parabolic, power, exponential, logarithmic, and hyperbolic. A method based on minimizing the ex ante forecast error was applied.
By changing the number of the last L years included in the prediction process, the minimum value of the ex ante error is sought.
For the optimal number of years, the forecast value is determined. The method was discussed using the example of predicting the values of selected motorization level indicators: the number of registered motor vehicles and passenger cars for selected cities in Poland: Gdańsk, Gliwice, Kraków, and Szczecin. Calculations were performed for data for the years (…)
The article contains 15838 characters.
Source: Logistyka Magazine