摘要
各单项预测方法权重的确定是区间型组合预测中的核心问题。与常规区间型组合预测方法相比,文章从熵值的角度出发,提出了三种简易可行的区间型组合预测权重确定的方法:将区间数转换成实数,利用相对熵确定区间型组合预测的权重;对于区间数自身,利用Shannon熵确定区间型组合预测的权重;将区间数转换成等价的二元联系数,利用联系熵确定区间型组合预测的权重。实例分析结果表明,所提出的三种权重确定方法简单可行,计算结果显示相应权重的区间型组合预测模型具有较高的预测精度。
The weight determination of each single forecasting method is the core of interval combination forecasting. Compared with conventional interval combination forecasting methods, this paper proposes three simple and feasible methods to determine the weight of interval combination forecasting from the perspective of entropy. Firstly, the interval number is converted into real number, and the relative entropy is used to determine the weight of the interval combination forecasting. Secondly, for the interval number itself, Shannon entropy is used to determine the weight of the interval combination forecasting. Thirdly, the interval number is converted into the equivalent binary relation number, and the relation entropy is used to determine the weight of the interval combination forecasting. The results of example analysis show that the three weight determination methods proposed are simple and feasible, and the calculation results show that the interval combination prediction model with corresponding weights has higher prediction accuracy.
作者
胡凌云
袁宏俊
李东勤
Hu Lingyun;Yuan Hongjun;Li Dongqin(School of Management Science and Engineering,Anhui University of Finance and Economics,Benghu Anhui 233030,China;School of Statistics and Applied Mathematics,Anhui University of Finance and Economics,Benghu Anhui 233030,China;School of Statistics,Donghei University of Finance and Economics,Dalian Liaoning 116025,China)
出处
《统计与决策》
CSSCI
北大核心
2021年第11期34-38,共5页
Statistics & Decision
基金
安徽省教育厅人文社会科学重点研究项目(SK2018A0431,SK2020A0028)
安徽省教育厅自然科学重点研究项目(KJ2020A0817)
安徽财经大学科学研究基金资助项目(ACKYC19038)。
关键词
区间组合预测
Shannon熵
相对熵
联系熵
interval combination forecasting
Shannon entropy
relative entropy
connexion entropy
