首页 | 本学科首页   官方微博 | 高级检索  
     

回归系统广义岭型主成分两步估计的优良性质
引用本文:黄云腾,朱宁,赵肖肖,黄敦明.回归系统广义岭型主成分两步估计的优良性质[J].桂林电子科技大学学报,2012(3):244-248.
作者姓名:黄云腾  朱宁  赵肖肖  黄敦明
作者单位:桂林电子科技大学 数学与计算科学学院,广西 桂林 541004
基金项目:国家自然科学基金(71001015)
摘    要:针对两步估计问题,在设计阵成病态的情形下,探讨了残差变量相关的多个方程回归系统如何转化为一般线性模型;与有偏估计相结合,得到该方程系统未知参数的估计模型———m 个相关方程回归系统的广义岭型主成分有偏两步估计;证明了该估计在相对效率的意义上比一般的两步估计、主成分两步估计、岭型主成分两步估计的效率高.

关 键 词:有偏两步估计  相关方程回归系统  相对效率

The good nature of two-step estimation of generalized ridge and principal components of regression system
Huang Yunteng,Zhu Ning,Zhao Xiaoxiao,Huang Dunming.The good nature of two-step estimation of generalized ridge and principal components of regression system[J].Journal of Guilin Institute of Electronic Technology,2012(3):244-248.
Authors:Huang Yunteng  Zhu Ning  Zhao Xiaoxiao  Huang Dunming
Affiliation:(School of Mathematics and Computational Science,Guilin University of Electronic Technology,Guilin 541004,China)
Abstract:Aiming at two-step estimation,the way of the multiple equation regression system transposing to a general linear model is discussed.Combining with the biased two-step estimation,a model for the estimation of unknown parameters of the regression system is obtained when its design matrix is pathological.It is proved that the biased two-step estimation is more efficient in the sense of the relative efficiency compared with the general two-step estimation,the two-step estimation of the principal component and the two-step estimation of ridge and principal components.
Keywords:biased two-step estimation  regression system of related equation  relative efficiency
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号