| Advertisement |
When forecast errors are squared and accumulated, give a measure of the discrepancy between the forecasts and the actual sales. Similarly, other values of 0.2, 0.3, α= etc. may be tried. The value that gives the least squared diviation of error would be optimum α. In practice, it is not infrequent to assign the value of a from experience. Where the computer activities are available, the above method of least squared deviations may be carried out for groups of items depicting similar sales behaviour.
Limitation of Exponential Smoothing
(i) The method is useful for short-term forecasting only.
(i) The method is useful for short-term forecasting only.
(ii) It relies solely on the past history of sales. There are cases where subjective estimates may provide better forecasts. There have been attempt to complement the two; forecasts from Exponential smoothing and subjective estimates, successful to a great extent, but not fully satisfactory.
