回归系数估计的渐近分析

回归系数估计的渐近分析 Asymptotic Analysis for an Estimate of Regression Coefficients 蔡新民 CAI Xin\|min 1 1 2 3 4 5 6 7 first-author 黄养新 HUANG Yang\|xin 2 1 2 3 4 5 6 7 武汉理工大学理学院 武汉理工大学理学院 School of Sciences,Wuhan University of Technology School of Sciences,Wuhan University of Technology Frontier Science&Technology Research Foundation,湖北,武汉,430070 Frontier Science&Technology Research Foundation,湖北,武汉,430070 Frontier Science & Technology Research Foundation, Harvard, 1244 Boylston Street, Suite 303, Chestnut Hill Frontier Science & Technology Research Foundation, Harvard, 1244 Boylston Street, Suite 303, Chestnut Hill Harvard Harvard Boylston Street,1244 Boylston Street,1244 Suite,303 Suite,303 Chestnut Hill Chestnut Hill ,USA ,USA 讨论正态线性模型回归系数的一种非线性有偏估计 ,给出了它的偏差及其均方误差的渐近展开式 .在均方误差意义下 ,得到了该估计优于最小二乘估计的渐近充要条件 A biased estimate of regression coefficients of a linear normal model is discussed. Its asymptotic expansions of bias and mean square error are derived and the asymptotic necessary and sufficient conditions are also shown for this estimate to dominate the best linear unbiased estimate. 线性模型 均方误差 渐近分析 linear model mean square error asymptotic analysis O212.1 国家科技部技术创新基金资助项目 （0 2C2 62 1 4 2 0 0 2 1 8） 2003-01-01 2021-04-01 1