基于Theil不等系数的IOWHA算子预测模型的性质研究

Study on Properties of Induced Ordered Weighted Harmonic Averaging(IOWHA) Operator Based on Theil Coefficients

为了克服传统的单项预测方法选取固定参数时的不足,在诱导有序加权调和平均算子(IOWHA算子)的基础上引入Theil不等系数,构建了基于Theil不等系数的IOWHA算子的最优组合预测模型.对该模型的预测精度、优性及非劣性进行定义,从理论角度探究了其非劣性组合预测、优性组合预测存在的充分条件.实例分析表明,该组合预测模型优于传统的组合预测模型,能够充分利用各个单项预测方法的信息并能提高预测精度,是一种优性组合预测.

In order to overcome the single forecastig method to take the fixed weight coefficient of defects,based on induced ordered weighted harmonic averaging operator(IOWHA operator),the optimal combination forecasting model of IOWHA operator based on Theil coefficient is introduced.The prediction accuracy,optimality and non-inferiority of the model are defined for the combined forecasting model.The sufficient conditions for the existence of non-inferiority combination forecasting and superior combination forecasting are studied from the perspective of theory.The example analysis shows that the combined forecasting model is better than the traditional combined forecasting model,can fully use the information of individual prediction methods and improve the prediction accuracy,it is a superior combination prediction.