一类位置不变的Pickands型估计量  

A Location Invariant Pickands-Type Estimator

作　　者：陶宝[1] 

机构地区：[1]重庆工商大学数学与统计学院,重庆400067

出　　处：《应用概率统计》2009年第5期449-460,共12页Chinese Journal of Applied Probability and Statistics

基　　金：国家自然科学基金(70371061);重庆市教委科学技术研究项目(030705)资助

摘　　要：当极值指标大于0时,本文提出了一种位置不变的Pickands型估计量,证明了该估计量的强弱相合性,给出了其渐近展式和强收敛速度,并对k2的最优选择进行了讨论,最后利用自适应性方法对该估计量和其它Pickands型估计量进行随机模拟分析,比较该估计量的优越性.In this paper, the author proposes a location invariant Pickands-type estimator as the extreme index is positive. Its weak and strong convergence are proved. Asymptotical representation and strong convergence rate are derived, and the optimal choice of k2 is found. At last simulation and comparison of this new kind of Pickands-type estimator with other known Pickands-type estimators are considered by using the adaptive algorithm.

关 键 词：弱相合性 强相合性 渐近展式 强收敛速度 最优选择 

分 类 号：O211.4[理学—概率论与数理统计]