摘要
传统的组合预测方法通常是针对非区间数据在最优准则下来建立最优预测模型的,但实际,中在不确定环境下样本数据和预测值序列均以区间数形式给出,因此有必要研究区间组合预测模型。文章引进连续区间诱导广义有序加权对数平均算子(C-IGOWLA)的概念,以Theil不等系数作为最优准则,提出了一种新的基于Theil不等系数的C-IGOWLA算子的区间组合预测模型,并结合实例验证了模型的有效性。
Traditional combined forecasting methods are often aimed at interval data under optimal rules to build optimal prediction models, but in practice, under uncertain environment, the sampled data and the series of predicted values are given in the form of interval numbers, hence the need to study interval combined forecasting models. This paper introduces the concept of Continuous Induced Generalized Ordered Weighted Logarithmic Averaging （C-IGOWLA）, and takes Theit coefficient as the optimal criterion to come up with a new interval combined forecasting model of the operator of C-IGOWLA based on Theil coefficient. Finally the paper uses examples to verifies the effectiveness of the proposed model.
出处
《统计与决策》
CSSCI
北大核心
2017年第15期32-35,共4页
Statistics & Decision
基金
教育部人文社会科学研究青年基金项目(12YJC630277)
安徽财经大学重点科研基金资助项目(ACKY1612ZDB)
关键词
C-IGOWLA
组合预测
区间数
Theil不等系数
continuous induced generalized ordered weighted logarithmic averaging （C-IGOWLA）
combination foresting
interval numbers
Theil coefficient
