摘要
本文研究了线性模型中回归系数β的经验Bayes(EB)估计。利用密度函数及其偏导数的核估计构造出β的EB估计,在一定条件下,证明了β的EB估计的收敛速度任意接近于1。最后,我们给出了满足定理条件的实例。
In this paper, empirical Bayes(EB) estimator for coefficient of regression β in linear models was proposed. By using the kernel estimators of density function and its derivatives, we prove that the convergence rate of EB estimator for β can be arbitrarily close to I under the suitable conditions. Finally, an example was given.
关键词
回归系数
核估计
收敛速度
coefficient of regression
kernel estimator
empirical Bayes estimator
convergence rate
