Weight Combination Forecasting The Identification And Application

The different forecast method can provide the different useful information, if we drops some methods which have the higher forecast deviations simply, it Will be able to lose some useful information. One scientific method carries on the combination forecast. The combination forecast can use each kind of forecast sample information the great limit, it's more systematic than comprehensively the single forecast model consideration question, can in the effective reduction single forecast model some random factors influence, thus enhancement forecast precision. The combination forecast the law computation is simple, the precision is high, has the very good usability.This article utilizes the combination forecast model to carry on the comparison, then to combines the forecast the weight to carry on the synthesis through several tax power method. Then to the forecast value which obtains carries on the contrast, elaborated with emphasis induces the order weighing method the combination to forecast, and unified the example has discussed different combination forecast weight determination in reality application to each kind of forecast model and the combination forecast model, forecast the result is steady, moreover the forecast result also showed this combination forecast model forecast precision often must be higher than the single item forecast model and other weighing method combination forecast value.The this article essential target is analyzes the combination forecast model the weight definite method, compared with each kind of forecast theory precision, to combines the forecast by this the weighing method to carry on the understanding and the application. Obtains through the example should selects the effective reliable forecast method according to the forecast different request and the characteristic and the applicable scope, to forecast the obtained result judges, Promptly revises and eliminates the improper model; the combination forecast the precision usually surpasses any list forecast the precision, the model integer increase may enhance the combination forecast the precision; introduced the M dimension induction order weighted average operator constitution in the M dimension order weighted average this kind of combination forecast foundation. With the order weighing method to the Chinese population forecast carries on the application, use the data about some cities for eight month-long with the induction order method of weighting. It has specifically demonstrated the induction order weighted average operator model establishment and the operation, the forecast effect is higher than the general methods.