右删失混合相依模型下几种危险率函数估计的渐近性质

Asymptotic Behavior of Several Kernel Estimators of Hazard Rate Function for Censored Dependent Data

摘要: 本文在右删失混合相依模型下,证明了密度函数和危险率函数的核类型估计具有与i.i.d.情形一样的渐近行为.并在极小平均平方误差准则下,得到了核危险率估计的最优窗宽为O(n^-1/5).

Abstract: Based on censored dependent data, we show that the kernel density and hazard rate estimator have exactly the same asymptotic performance as in the i.i.d. case, and the proper bandwidth for the kernel estimator of hazard rate λ(t) should be of the order O(n^-1/5).