本文构造了竞争风险场合分布函数的乘积极限(PL)型估计,运用经验过程的强逼近理论及Toylor展开方法,给出了PL型估计在全直线上的强一致收敛速度及其充分必要条件。
In this paper, we introduce the product-limit estimator of a distribution function under competing risks case. We give the rate of strong uniform convergence of the PL-type estimator over the whole line and drive sufficient and necessary conditions by the strong approximation of empirical process and Taylor expansion.