随机右删失部分线性模型中小波估计的强相合性

摘要

研究随机右删失部分线性模型中回归参数β和回归函数g(t)的估计问题.分别在删失分布函数已知和未知的情形下,基于变换后的观测数据构造了β和g(t)的小波估计,在适当的条件下,证明了它们均具有强相合性.%The problem of estimating the regression parameter β and the regression function g(t) of randomly right censored partially linear models were studied. Under the situations that the censored distribution function is known and unknown,respectively,the wavelet estimators of β and g(t) are constructed based on the transformed observations. Under appropriate conditions, it has been proved that both estimators possess strong consistency.