基于区间数距离和ICOWA算子的变权区间组合预测模型

The Variable Weight Interval Combination Forecasting Model Based on Distance between Interval-valued Fuzzy Sets and ICOWA Operator

针对预测值为区间数形式,将区间数距离和诱导连续有序加权平均算子ICOWA结合,以ICOWA算子下实际预测序列与组合预测序列的距离最小为准则,构建了变权组合预测模型,并证明了模型的非劣性。最后,通过实例说明了模型能够有效提高预测精度。

A method of calculating the variable weight coefficient of the comb-ination forecasting is proposed when forecasting value series are given in the form of interval numbers.For the prediction value is in the form of interval number,the interval number distance is combined with the induced continuous ordered weighted average operator ICOWA.Based on the minimum distance between the actual prediction sequence and the combination prediction sequence under ICOWA operator,the variable weight combination prediction model is constructed,and the non inferiority of the model is proved.Finally,an example shows that the model can effectively improve the prediction accuracy.