Comparative analysis of accuracy of selected methods of building of combined forecasts and meta-forecast
Languages of publication
In this paper the author presents a method of building a meta-forecast as an arithmetic mean of the combined forecasts set by various methods. The empirical example, in which the forecasts (individual, combined and meta-forecasts) are determined for the microeconomic variable with seasonal fluctuations, is the illustration of theoretical considerations. The accuracy of meta-forecasts is compared with the accuracy of their component combined forecasts and individual forecasts. The empirical studies confirm the usefulness of meta-forecasts. In most cases, they have lower errors than their component combined forecasts, also they are more accurate than individual forecasts.
- Aksu C., Gunter S. (1997), The usefulness of heuristic N(E)RLS algorithms for combining forecasts, Journal of Forecasting 16: 439–463.
- Armstrong J. (2001), Principles of Forecasting: a Handbook for Researchers and Practitioners, Kluwer Academic Publishers, Boston.
- Bates J., Granger C. (1969), The combination of forecasts, Operational Research Quarterly 20.
- Granger C., Newbold P. (1974), Experience with forecasting univariate time series and the combination of forecasts, Journal of the Royal Statistical Society, A, 137.
- Kaźmierska-Zatoń M., Zatoń W. (2010), Multi-criteria combined forecasts, Econometrics 28 (91), 59–75, Wrocław.
- Liu B., Shi S., Xu L. (1996), Improving the accuracy of nonlinear combined forecasting using neural networks, Expert Systems with Applications 16: 49–54.
- Perzyńska J. (2010), Budowa prognoz kombinowanych z wykorzystaniem sztucznych sieci neuronowych, [in:] P. Dittmann, E. Szabela-Pasierbińska (eds.), Prognozowanie w zarządzaniu firmą, Prace Naukowe Uniwersytetu Ekonomicznego we Wrocławiu nr 103, 133–145.
Publication order reference