聚集数据线性模型参数的广义聚集压缩双参数估计  

Compressed Aggregate Dual-parameter Estimation for Linear Model Parameters of Aggregate Data

作　　者：左卫兵 牛艳飞 ZUO Weibing;NIU Yanfei(College of Mathematics and Statistics, North China University of Water Resources andElectric Power, Zhengzhou 450046, China)

机构地区：[1]华北水利水电大学数学与统计学院,河南郑州450046

出　　处：《河南教育学院学报（自然科学版）》2018年第4期1-7,共7页Journal of Henan Institute of Education(Natural Science Edition)

基　　金：河南省基础与前沿技术研究计划(152300410112)

摘　　要：主要讨论聚集数据线性模型的参数估计问题.在广义聚集双参数(GADP)估计的基础上利用协方差改进法提出了改进广义聚集双参数(IGADP)估计.然后又结合IGADP估计和Stein估计,提出了广义聚集压缩双参数(GACDP)估计.此外,在均方误差准则下,导出了GACDP估计优于IGADP估计、广义聚集双参数(GADP)估计、改进广义岭(IGR)估计及改进P-K(IPK)估计的充分条件.最后通过数值模拟分析验证了所得理论结果.Mainly discusses the parameter estimation problem of the linear model of aggregated data. An improved generalized aggregation double parameter (IGADP) estimation is proposed based on the covariance improvement method based on the generalized aggregated double parameter (GADP) estimation. In addition, combine the IGADP estimation and the Stein estimation to propose the generalized aggregated compression double parameter (GACDP) estimation. Under the mean square error criterion, the sufficient conditions for the GACDP estimation to be superior to the IGADP estimation, the generalized aggregation double parameter (GADP) estimation, the improved generalized ridge (IGR) estimation and the improved P-K (IPK) estimation are derived. Finally, the theoretical results are verified by numerical simulation.

关 键 词：聚集数据线性模型 GACDP估计 IGADP估计 均方误差 数值模拟 

分 类 号：O212.1[理学—概率论与数理统计]