摘要
本文考虑多元极值分布的参数估计，给出了分步估计渐近协差阵的近似表示，并对维数P=2，5及相关参数α=0．001，0．01；0．1（0．2），0．9；0．99，0．999的各种组合，计算了分步估计关于最大似然估计的渐近效率，分析了各种参数及维数对渐近效率的影响．分步估计是一种合理、简单而且有较强实用意义的估计方法．
The parametric estimations of multivariate extreme value distribution (in Logistic model) are considered. The approximate formulae for asymptotic covariance matrix of stepwise method are indicated, and asymptotic relative efficiencies of the stepwise method with regard to the maximum likelihood estimation are computed for certain combinations of dimension p= 2, 5 and dependent parameter α = 0.001, 0.01, 0.1 (0.2), 0.9, 0.99, 0.999. The effect on asymptotic efficiencies by all kinds of parameter and dimension are considered. We point out that the stePWise estimation is a reasonable, simple and widely applicable method.
出处
《系统科学与数学》
CSCD
北大核心
1997年第3期244-251,共8页
Journal of Systems Science and Mathematical Sciences
基金
国家自然科学基金
关键词
最大似然估计
分步估计
参数估计
多元极值分布
Asymptotic efficiency, generalized extreme value distribution, Gumbel distribution, maximum likelihood estimation, multivariate extreme value distribution
