×

Initial conditions estimation for improving forecast accuracy in exponential smoothing. (English) Zbl 1284.62586

Summary: We analyze the importance of initial conditions in exponential smoothing models on forecast errors and prediction intervals. We work with certain exponential smoothing models, namely Holt’s additive linear and Gardner’s damped trend. We study some probability properties of those models, showing the influence of the initial conditions on the forecast, which highlights the importance of obtaining accurate estimates of initial conditions. Using the linear heteroscedastic modeling approach, we show how to obtain the joint estimation of initial conditions and smoothing parameters through maximum likelihood via box-constrained nonlinear optimization. Point-wise forecasts of future values and prediction intervals are computed under normality assumptions on the stochastic component. We also propose an alternative formulation of prediction intervals in order to obtain an empirical coverage closer to their nominal values; that formulation adds an additional term to the standard formulas for the estimation of the error variance. We illustrate the proposed approach by using the yearly data time-series from the M3-Competition.

MSC:

62M20 Inference from stochastic processes and prediction
62P30 Applications of statistics in engineering and industry; control charts

Software:

LBFGS-B; expsmooth
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] Arnold JRT, Chapman SN, Clive LM (2008) Introduction to materials management, 6th edn. Pearson Prentice Hall, Upper Saddle River
[2] Bermúdez JD, Segura JV, Vercher E (2006a) Improving demand forecasting accuracy using non-linear programming software. J Oper Res Soc 57:94–100 · Zbl 1121.90301
[3] Bermúdez JD, Segura JV, Vercher E (2006b) A decision support system methodology for forecasting of time series based on soft computing. Comput Stat Data Anal 51:177–191 · Zbl 1157.62502
[4] Bermúdez JD, Segura JV, Vercher E (2007) Holt–Winters forecasting: an alternative formulation applied to UK air passenger data. J Appl Stat 34:1075–1090
[5] Bermúdez JD, Segura JV, Vercher E (2008) SIOPRED: a prediction and optimisation integrated system for demand. Top 16:258–271 · Zbl 1154.62066
[6] Bermúdez JD, Segura JV, Vercher E (2010) Bayesian forecasting with the Holt–Winters modes. J Oper Res Soc 61:164–171 · Zbl 1193.90016
[7] Byrd RH, Lu P, Nocedal J, Zhu C (1995) A limited memory algorithm for bound constrained optimization. SIAM J Sci Comput 16:1190–1208 · Zbl 0836.65080
[8] Corberán-Vallet A, Bermúdez JD, Vercher E (2011) Forecasting correlated time series with exponential smoothing models. Int J Forecast 27:252–265
[9] Fildes R, Kingsman B (2011) Incorporating demand uncertainty and forecast error in supply chain planning models. J Oper Res Soc 62:483–500
[10] Fildes R, Nikolopoulos K, Crone SF, Syntetos AA (2008) Forecasting and operational research: a review. J Oper Res Soc 59:1150–1172 · Zbl 1153.90009
[11] Gardner ES Jr (2006) Exponential smoothing: the state of the art–part II. Int J Forecast 22:637–666
[12] Gardner ESJr, McKenzie E (1985) Forecasting trends in time series. Manag Sci 31:1237–1246 · Zbl 0617.62105
[13] Harvey A (1989) Forecasting, structural time series models and the Kalman filter. Cambridge University Press, Cambridge
[14] Hyndman RJ, Koehler AB (2006) Another look at measures of forecast accuracy. Int J Forecast 22:679–688
[15] Hyndman RJ, Koehler AB, Ord JK, Snyder RD (2005) Prediction intervals for exponential smoothing using two new classes of state space models. J Forecast 24:17–37
[16] Hyndman RJ, Koehler AB, Ord JK, Snyder RD (2008) Forecasting with exponential smoothing. The state space approach. Springer, Berlin
[17] Hyndman RJ, Koehler AB, Snyder RD, Grose S (2002) A state space framework for automatic forecasting using exponential smoothing methods. Int J Forecast 28:439–454
[18] McKenzie E, Gardner ES Jr (2010) Damped trend exponential smoothing:a modelling viewpoint. Int J Forecast 26:661–665
[19] Makridakis S, Hibon M (1991) Exponential smoothing: The effect of initial values and loss functions on post-sample accuracy. Int J Forecast 7:317–330
[20] Makridakis S, Hibon M (2000) The M3-competition: results, conclusions and implications. Int J Forecast 16:451–476
[21] Ord JK, Koehler AB, Snyder RD (1997) Estimation and prediction for a class of dynamic nonlinear statistical models. J Am Stat Assoc 92:1621–1629 · Zbl 0912.62104
[22] Segura JV, Vercher E (2001) A spreadsheet modeling approach to the Holt–Winters optimal forecasting. Eur J Oper Res 178:154–167 · Zbl 0991.90086
[23] Syntetos AA, Boylan JE, Disney SM (2009) Forecasting for inventory planning: a 50-year review. J Oper Res Soc 60:S150–S160 · Zbl 1168.90305
[24] Taylor JW (2003) Exponential smoothing with a damped multiplicative trend. Int J Forecast 19:715–725
[25] Tratar LF (2010) Joint optimisation of demand forecasting and stock control parameters. Int J Prod Econ 127:173–179
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.