SIOPRED: a prediction and optimisation integrated system for demand
- 82 Downloads
This paper presents a forecasting support system based on the generalised Holt-Winters exponential smoothing scheme to forecast time series of levels of demand. It is conceived as an integrated tool which has been implemented in Visual Basic. For improving the accuracy of automatic forecasting it uses an optimisation-based scheme which unifies the stages of estimation of the parameters and model selection. Based on this scheme, suitable forecasts and prediction intervals are obtained. The performance of the proposed system is compared with a number of well-established automatic forecasting procedures with respect to the 3003 time series included in the M3-competition.
KeywordsForecasting Holt-Winters method Non-linear optimisation Decision support systems
Mathematics Subject Classification (2000)62M10 62M20 62P30 90C30
Unable to display preview. Download preview PDF.
- Bermúdez JD, Segura JV, Vercher E (2007a) Predicción del consumo de agua con SIOPRED. In: Rojas I, Pomares H (eds) Actas de simposio de inteligencia computacional SICO’2007. Thomson, Madrid, pp 447–451 Google Scholar
- Box GEP, Jenkins GM, Reinsel GC (1994) Time series analysis, forecasting and control, 3rd edn. Prentice Hall, Englewood Cliffs Google Scholar
- Segura JV, Vercher E (2002) A mathematical programming model for the Holt-Winters multiplicative method. Working paper I-2002-02, Centro de Investigación Operativa Google Scholar
- Tashman LJ, Hoover J (2001) Diffusion of forecasting principles through software. In: Armstrong JS (ed) Principles of forecasting. Kluwer Academic, Boston, pp 651–676 Google Scholar
- Theil H (1966) Applied economic forecasting. North-Holland, Amsterdam Google Scholar
- Winters PR (1960) Forecasting sales by exponentially weighted moving averages. Manag Sci 6:324–342 Google Scholar
- Yu X, Kacprzyk J (eds) (2003) Applied decision support with soft computing, vol 124. Springer, Berlin Google Scholar