基于指数支撑度的IGOWLA算子的最优组合预测模型  

The Combination Forecasting Model Based on Exponential Supporting Degree and Induced Geometric Ordered Weighted Logarithm Averaging(IGOWLA) Operator

作　　者：孙浩[1] 杨桂元[1] SUN Hao;YANG Gui-yuan(School of Statistics and Applied Maths,Anhui University of Finance and Economics,Bengbu,Anhui 233030,China)

机构地区：[1]安徽财经大学统计与应用数学学院,安徽蚌埠233030

出　　处：《南华大学学报（自然科学版）》2018年第1期58-62,共5页Journal of University of South China：Science and Technology

基　　金：国家社会科学基金研究项目资助(12BTJ008);安徽财经大学研究生科研创新基金项目(ACYC2016113)

摘　　要：在广义诱导有序加权对数平均算子(IGOWLA算子)的基础上,引入指数支撑度以及构建了一种基于支撑度的IGOWLA算子的最优组合预测模型,并对该模型的预测精度、优性及非劣性给出定义.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.Based on the induced geometric ordered weighted logarithm averaging operator(IGOWLA operator),the paper construct the optimal combination forecasting model of IGOWLA operator on the basis of an exponential supporting degree and power error are introduced.Then,the prediction accuracy,superiority and non-inferiority of this model are defined.Finally,an example was illustrated by using the model which can make full use of the information from the single forecasting method.The forecasting precision is superior to the traditional single forecasting model as well.It is concluded that the method is a superior combination forecasting method.

关 键 词：指数支撑度 组合预测 IGOWLA算子 

分 类 号：F224.0[经济管理—国民经济]