随机删失场合半参数回归模型参数估计的重对数律‘

本文探讨了随机删失场合半参数回归模型的参数估计问题.考虑半参数回归模型Y =X}}3 + g(T)十。，其中(X,T)’为取值于Kp X 0,1〕上的随机向量，月为1'维未知参数向量，8为定义于【0.1]上的未知函数，。为随机误差，Ee = 0 . Eez = az }。未知，且(X ,T)与。独立，).被一个与之独立的随机变量V所截.此时仅能观察到:Z=min(Y,V),o=1(Y簇V)，参数I3,az的估计量禽及公 z可综合非参数的权函数估计法与参数的最小二乘估计方法得到.本文对核函数的情形得到了念及ar z的精确收敛速度即重对数律.

The Law of Iterated Logarithm of Estimation for Parameter in a Semiparametric Regression Model with Censored Data

This paper studied the parameter estimation in a semiparametric regression model with censored da-ta. A semiparametric regression model Y=X'-f-g (T ) +ewas taken into account,where (X,T) is a random vector onm R0X 0,1}，is a p-dimensional unknownparameter ,g·)is an unknown function on 0, l},e is a random error with mean 0 and variance az, (X，T) is independent of e. And Y may be censored and therefore not completely observable,where the ten-soring variable V and the variable Y are independent mutually, We estimate and az by nonparametricweight function and least square method. As a result ,this paper prove that these estimationsandzoand a2 have the law of iterated logarithm.