The meta-elliptical distributions with given marginals

被引:240
|
作者
Fang, HB
Fang, KT [1 ]
Kotz, S
机构
[1] Hong Kong Baptist Univ, Dept Math, Kowloon, Hong Kong, Peoples R China
[2] George Washington Univ, Washington, DC 20052 USA
关键词
conditional quantile; copulas; elliptically contoured distributions; Kendall's tau; likelihood ratio dependence; multivariate distribution; regression dependence;
D O I
10.1006/jmva.2001.2017
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
Based on an analysis of copulas of elliptically contoured distributions, joint densities of continuous variables with given strictly increasing marginal distributions are constructed. A method utilized for this procedure is to embed the spherical distribution quantile transformation of each variable into an elliptically contoured distribution. The new class of distributions is then called meta-elliptical distributions. The corresponding analytic forms of the density, conditional distribution functions, and dependence properties are derived. This new class of distributions has the same Kendall's rank correlation coefficient as meta-Gaussian distributions. As an extension of elliptically contoured distributions, some new classes of distributions are also obtained. (C) 2002 Elsevier Science (USA).
引用
收藏
页码:1 / 16
页数:16
相关论文
共 50 条