You achieve better results than those received from the moving average model by introducing weighting factors for each past value. This means that every past value is weighted with the factor R. The sum of the weighting factors is 1 (see formulae (3) and (4) below).
If the time series to be forecast contains trend-like variations you will receive better results by using the weighted moving average model rather than the moving average model. The reason for this is that more weight is given to more recent data when determining the average than to older data, that is, if you selected appropriate weighting factors. Therefore, the system will be able to react more quickly to a change in level.
This model, however, depends strongly on your choice of weighting factors. If the time series pattern changes, you must also adapt the weighting factors.
No comments:
Post a Comment