线性模型误差方差的经验似然估计

Empirical Likelihood Estimators of the Error Variances in Linear Models

摘要: 本文应用经验似然方法得到了线性模型误差方差的一类新的估计,证明了估计的渐近分布为正态分布且渐近方差不超过常用的误差方差估计的渐近方差,同时给出了渐近方差的显式表达.

Abstract: We apply the empirical likelihood technique to propose a new class of estimators of the error variance in linear models. It is shown that the proposed estimators are asymptotically normally distributed with asymptotic variances not greater than that of the usual estimators of the error variance. And the closed forms of the asymptotic variances of the estimators are presented.