The Optimal Combined Forecasting Model and Application under the Two Kinds of Criterions based on IOWA Operator
Qimin Chen · 2013
In the traditional combination forecasting,the weights are generally fixed in each single forecasting model.In fact,the forecasting accuracy is always changeable with time.Accordingly,the weight coefficient of single forecasting model should be changed.Based on the concept of forecasting effective measure and induced ordered weighted averaging(IOWA) operator,the effective degree and sum of error squares are regarded as criterias to a criteria to the combination forecast models.Finally,we bring up the application of the models,which shows that the new model is a kind of superior combination forecasting.