摘要
考虑半参数回归模型Yi=xiβ+g(ti)+iσiε,i=1,2,…,n,其中2iσ=f(ui).当Yi因受某种随机干扰而被随机右删失时,就删失分布未知的情形,利用所获得的删失数据定义了β与g(t)的估计^βn和^gn(t),在适当的条件下,证明了^nβ的渐近正态性,同时得到了^gn(t)的最优收敛速度.
It was considered that semiparametric regression model Yi=xiβ＋g（ti）＋σiEi,i=1,2,…,n, where σ^2i =f（ui）. When Yi is randomly censored on the right because of certain disturbance, under the situation that the censored distribution function is unknown, the estimators of parameter β and regression function g（t） are constructed based on censored observations. Under appropriate conditions, it has proved that estimater β^-n, has asymptotic normality and g^-n（t） has optimal convergence rate.
出处
《纺织高校基础科学学报》
CAS
2008年第2期223-229,共7页
Basic Sciences Journal of Textile Universities
关键词
随机删失
半参数回归模型
渐近正态性
最优收敛速度
random censorship
semiparametric regression model
asymptotic normality
optimal convergence rate
