条件中位数的随机窗宽核估计  

Random Window-Width Kernel Estimates of Conditional Median

作　　者：何仲洛 

机构地区：[1]湖州师专数学系

出　　处：《湖州师专学报》1992年第5期1-22,共22页

摘　　要：本文涉及到条件中位数的随机窗宽核估计,在适当的正则性条件下,我们在定理1中证明了条件中位数随机窗宽核估计ξ_(x n)的渐近正态性,我们也考虑了ξ_(xn)的渐进正态性的收敛速度,定理2提供了速度为O(n^(-1/(d+2))logn),对于ξ_(xn)的Bahader表示被给出在定理3中,这个表示理论的一个涉及到ξ_(x n)的重对数律的应用则作出在定理4中.This paper is concerned with random window-width kernel estimates of conditional median. Under appropriate regularity conditions , we prove asymptotic normality of random window-width kernel estimators ξxn of the conditional median in theorem 1 . We also consider the rate of convergence in connection with the asymptotic normality of ξxn , theorem 2 providesthe rateO(n d+2 log n ) . Bahader's representation for ξxn, is presented intheorem 3 . An application of the representation theory will be made in theorem 4. in connection with the law of the iterated logarithm for ξxn.

关 键 词：条件中位数 随机窗宽 核估计 

分 类 号：O212.1[理学—概率论与数理统计]