随机右删失下核光滑乘积限估计的大样本性质

The large sample property of kernel smoothing product limit estimation under random right censorship

讨论了随机右删失下乘积限估计Fn(t) 与核光滑乘积限估计Fn(t) 之差(Fn - Fn)(t)的大样本性质,得出了Fn(t)的一般大样本性质,特别是n1/2(F(t)- F(t))= γn(t)→w Gp(t)的充分与必要条件.

This paper studies the large sample property of |F n(t)- n(t)| and obtains the large sample property of kernel smoothing product limit estimation  n(t) under random right censorship, especially the necessary and sufficient condition implying n 1/2 ((t)-F(t))= n(t)→wG p(t).