在医学和金融保险等领域中,研究者通常关注的是竞争风险下失效时间的分布特征,特别是分布的尾部性质.本文对竞争风险下的剩余寿命分位数的估计提出了一种光滑非参数的方法,解决了直接基于经验分布估计可能无解的问题.同时,建立了所提出估计的渐近性质.并通过数值模拟说明了在均方误差的意义下,光滑估计比非光滑估计更有效.
In many fields, such as medicine and finance and insurance, researchers are often interested in the distribution of failure time under competing risks, especially tail behavior of the distribution. In this article, a smoothed nonparametrie estimator of quantile residual lifetime under competing risks is proposed, which ensures that the numerical solution does exist. The asymptotic properties of the smoothed estimator are established. Also, simulation studies are conducted and show that the smoothed estimate is more efficient by means of deficiency.