Asymptotic properties of estimators in semiparametric regression model under random censorship;

In this paper the convergence rates of the improved kernel regression estimate and the kernel regression estimate are obtained under random censorship.

This article discusses the parameter estimation problem in the two-parameter exponential distribution based on complete sample, type I censoring sample and random censorship sample, respectively.

Based on the iteration algorithm which is proposed to deal with non-random censoring data in past paper,the data-filling method is modified so that the virtual complete data obtained from the algorithm is much closer to the real complete data in the case of Weibull distribution under type I censoring data.

Based on the moment invariance criterion and corresponding algorithm which is proposed to deal with non-random censoring data, in this paper, the consistency of estimator is proved in the case of the Weibull distribution under type I censoring data, through controlling the quartile probability of the data filling algorithm.

The large sample property of kernel smoothing product limit estimation under random right censorship;

We propose a local quasi-likelihood weighted estimator for generalized semiparametric models when the Covariates are missing at random.

We develop two local quasi-likelihood imputation estimators for mean in a generalized varying-ciefticient model when response variables are missing at random.

We develop imputation estimators of mean of responses for semiparametric varying-coefficient model with response variables missing at random.