Censored Data In The Value Of The Regression Model

Median regression models with censored data as often-used models have been discussed in recent years, when the censoring distribution and the covariates are independent, it is often used the Kaplan-Meier estimator to estimate the censoring distribution, but when the censoring distribution depends on the covariates, in the third chapter, we propose semiparametric procedures for estimating parameters using the generalized product-limit estimator which based on weighted estimating equation, and derive its asymptotic properties. The estimator of the regression model is shown to be asymptotical normal distribution, improved estimating equation and efficiency are studied next, numerical studies are conducted to show estimator perform well with small samples and the resulting inference is reliable in circumstance of practical importance. Real data example is given. But what has discussed is restricted to the full censoring indicators, when there is missing censoring indicators, in the forth chapter, we propose semiparametric procedures for estimating parameters, we also get the estimating function, obtain the asymptotic properties, and give the improved estimating function, numerical study show that our estimators perform well. In the last chapter, we discuss when the censoring distribution are not the same, and when there are missing censoring indicators, we give the estimating method, discuss the asymptotic properties.