聚集数据线性模型参数的多元聚集改进的广义岭估计的相对效率  

The Relative Efficiency of the Improved Generalized Ridge Estimator for Multivariate Aggregation of the Parameters of the Linear Model with Aggregated Data

作　　者：丁洁 余新宏 张海波 DING Jie;YU Xin-hong;ZHANG Hai-bo(Basic Course Teaching Department,Hefei University of Economics,Hefei 230011,China)

机构地区：[1]合肥经济学院基础课教学部,安徽合肥230011

出　　处：《南京晓庄学院学报》2024年第6期96-101,共6页Journal of Nanjing Xiaozhuang University

基　　金：2022年度安徽省科研编制计划项目(2022AH052621);2023年度安徽省科研编制计划项目(2023AH052443)。

摘　　要：针对聚集数据的复共线性模型,岭估计在参数的稳定性方面过于强调,但参数的无偏性方面被忽视了。利用Stein压缩理论技术,鉴于广义Liu型估计方法之上,给出了多元聚集改进的广义岭估计。研究了该类改进估计的优良性,论证了其中两种相对效率的上界限问题。In this paper,based on Liu-type and Stein-type compression estimation method,an improved generalized ridge estimation of multivariate aggregation for ill-conditioned aggregated data model is proposed.The superiority of the improved generalized ridge estimator for multivariate aggregation is discussed,and the upper bounds of the two relative efficiencies are obtained.

关 键 词：聚集数据 多元聚集 广义岭估计 相对效率 

分 类 号：O212.4[理学—概率论与数理统计]