一类基于OWA算子的组合预测模型及其性质

A Kind of Combination Forecasting Model and Its Properties Based on Ordered Weighted Averaging(OWA) Operator

本文引进近年来提出的有序加权平均(OWA)算子,建立新的组合预测模型。首先指出最大绝对误差最小化的组合预测模型是其特例,并给出其线性规划的求解方法;然后对该模型提出非劣性组合预测、预测方法优超,冗余度等新概念;最后探讨了非劣性组合预测以及冗余预测方法的存在性,并给出冗余信息的判定定理。

Ordered weighted averaging （OWA） operators, which have been developed in recent years, are introduced to build the corresponding combination forecasting model in this paper. Firstly, it is pointed out that the combination forecasting model of minimizing maximum absolute error is the special case and the solution of this model is given by linear programming. Then some new concepts, such as noninferior combination forecasting dominant forecasting method, redundant degree etc, are put forward. Finally, we discusse the existence of noninferior combination forecasting and redundant forecasting method. The method of determining redundant information is also given in a theorem.