安徽大学学报（自然科学版）2025，Vol.49Issue(4)：18-27,10.DOI:10.3969/j.issn.1000-2162.2025.04.003
基于COWA算子的区间组合预测的均方误差分解及其应用
Mean square error decomposition for interval combination forecasting based on COWA operator and its application
摘要
Abstract
The concepts of interval upper and lower bound error covariance coefficients and interval inconsistency measures were proposed for the combined prediction method of weighted arithmetic averaging of the COWA operator on interval-type data.Using the COWA operator,we obtained expressions for the decomposition of the mean squared error metric of the combined forecast on interval data.The results showed that the combined forecast mean squared error was positively correlated with the mean squared error of each individual forecasting method and the coefficient of covariance of the upper and lower bound errors of the interval,and it was negatively correlated with the interval inconsistency measure of the individual forecasting methods.The mathematical properties of the decomposition of the mean squared error based on the COWA operator interval combined forecasts,as well as the individual forecasting methods,were discussed.To validate the decomposition framework,the factors proposed in this paper that affected the magnitude of the mean square error of the intervals were verified based on four single interval forecasts for two data sets,using different combinations of forecasts that gave the corresponding indicator values.This study provided scientific basis for the selection of single interval forecasting methods.关键词
区间预测/COWA算子/均方误差分解/不一致性测度Key words
interval forecasting/COWA operator/mean square error decomposition/inconsistency measure分类
数理科学引用本文复制引用
张琦,陈华友,余纾涵,汤天健..基于COWA算子的区间组合预测的均方误差分解及其应用[J].安徽大学学报（自然科学版）,2025,49(4):18-27,10.基金项目
国家自然科学基金资助项目(72371001) （72371001）
安徽大学创新训练项目(S202210357001) （S202210357001）
