New Approach for Determining the Smoothing Constant of a Single Exponential Smoothing Method

Full Text PDF PDF
Author(s) H.G. MU’AZU
Pages 717-727
Volume 3
Issue 11
Date November, 2014
Keywords Single exponential, Arbitrary, Heuristic, Smoothening, Forecasting

In this research work, a heuristic algorithm has been developed for deriving the exponential constant of the single exponential smoothing method. The approach bound the constant as . This is as against the arbitrary choice of the constant in the open interval as , which is the existing approach. Thus, the mathematical approach of determining the constant proposed in this research will guide analysts and users of single exponential smoothing model to derive same exponential constant for a given problem rather than the arbitrary choice which may results into different users or analysts choosing different exponential constants for the same problem. The determination of the constant proposed in this research is based on the number of historical data. The approach was tested empirically on thirty (30) documented problems and results were compared, graphically, with that obtained through the arbitrary choices of the exponential constants employed on same problems. The Exponential constants derived were found to be suitable for both smoothing time series data and for forecasting.

< Back to November Issue