Von—Mises统计量正则函数的自助(Boot—Strap)法

Bootstrapping the Functionals of Von-Mises Statisfics

本文基于近似独立的刀切虚拟值,对Von-Mises统计量的正则函数(?)建立了Bootstrap法。在适当的条件下证明了f(V_n)的Bootstrap逼近的相合性,并讨论了这种逼近的收敛速度,达到了理想的Berry-Esscen下限。

In this paper, based on the approximate indepedence of jackknife pseudo-values, we use the Boot-strap method to approximate the unknown distribution of Functionals of Von-Mises statistics. The consistency and accuracy of the bootstrapping approxiouation have been discussed.