协变量区间删失时线性模型的参数估计

Estimation of Parameters of a Linear Regression Model under Interval-censored Covariate

本文研究了当协变量为区间数据时的线性模型,通过构造区间数据变量的条件均值,得到了回归参数的估计,当协变量的分布已知时,证明了估计的无偏性与强相合性.对协变量的分布未知的情形也作了讨论.文中还作了若干模拟计算,从模拟的结果不难发现,利用本文提出的方法所获得的估计简便且具有较高的精度.

In this paper, we consider a linear regression model under interval-censored covariate. By constructing conditional mean of the interval censored covariate, the estimators of regression parameters are botained. The unbiasness and consistency of the estimators are proved when the distribution of covariate is given. We also discuss the situation when the distribution of covariate is not given. Some simulation results indicate that our method performs very good in terms of the accuracies of the estimation.