Combined Forecasting and Determining Weights Method

In order to get a better customer demand forecast result, we need to use two or more prediction methods for customer demand forecasts. Then the forecast results are combined according to weight value for every single result obtained by different prediction methods. By doing so maybe we can fully use obtained data information, thereby to improve forecast accuracy and get better prediction results. 

The key points of combined forecasting methods are the appropriate selection and weight determination of the single forecasting method. If the decision makers 

Set coj (i = 1,2. m) as ith method for the prediction of weight, then J2T=i o i= 1 Yit is obtained forecast value by using one of the prediction methods at time t, the forecasting results at time t can be expressed as  

The solution steps of m = (coi,cd2,. .oj ) by using fuzzy consistent matrix  

Mi and Mj represent the degree of relative importance of the prediction methods i and j given by decision makers. 

---