Some comments on the initialization of exponential smoothing
- 1 January 1984
- journal article
- research article
- Published by Wiley in Journal of Forecasting
- Vol. 3 (1) , 79-84
- https://doi.org/10.1002/for.3980030109
Abstract
It is shown that the traditional choice for the initial smoothed statistics in general exponential smoothing leads to the same forecasts as the equivalent ARIMA model, provided that one uses zero starting values for the initial shocks. In addition, an initialization which uses ‘backforecasts’ as initial smoothed statistics is considered, and its relationship to unconditional least squares is explored.Keywords
This publication has 12 references indexed in Scilit:
- Parsimony and Its Importance in Time Series ForecastingTechnometrics, 1981
- Finite sample properties of estimators for autoregressive moving average modelsJournal of Econometrics, 1980
- The likelihood function of stationary autoregressive-moving average modelsBiometrika, 1979
- Conditions for the optimality of exponential smoothing forecast proceduresMetrika, 1978
- A Monte Carlo study of autoregressive integrated moving average processesJournal of Econometrics, 1978
- An Analysis of General Exponential SmoothingOperations Research, 1976
- The Optimality of General-Order Exponential SmoothingOperations Research, 1974
- A Comparison of Standard Forecasting Systems with the Box-Jenkins ApproachJournal of the Royal Statistical Society: Series D (The Statistician), 1974
- The Fundamental Theorem of Exponential SmoothingOperations Research, 1961
- Optimal Properties of Exponentially Weighted ForecastsJournal of the American Statistical Association, 1960