文档介绍：基于连接函数的二元极值 Copula 的矩估计
黄超,林金官
东南大学数学系,南京 210096
摘要:本文研究了二元极值 Copula 的矩估计问题。通常极值 Copula 可以由一个 Pickands 相
依函数表示,因此本文基于连接函数针对 Pickands 相依函数提出一种二元极值 Copula 的矩
估计方法,该方法可以看作是现有非参数估计方法的一个推广。文章通过数值模拟给出了该方
法的有限样本性质。同时,本文还应用了参数方法和矩估计方法分析了两个不同位置的最大周
末车速数据。
关键词:二元极值 copula; Pickands 相依函数; 连接函数; 普通最小二乘法.
中图分类号:
Moments-type Estimation of Bivariate Extreme
Value Copulas Based on Link Functions
HUANG Chao , LIN Jin-guan
Department of Mathematics, Southeast University, Nanjing 210096
Abstract: This paper considers the moments-type estimation of bivariate extreme value
copulas. Usually the inference on an extreme value copula proceeds via its Pickands
dependence function. In this paper a new estimation method for the Pickands dependence
function is proposed based on link functions, which can be treated as an extension of the
existing nonparametric estimates. The finite sample performance is investigated by simulation
studies, and the maximum weekend car speed data registered at two given locations is
analyzed via parametric method and our proposed method.
Key words: Bivariate extreme value copula; Pickands dependence function; Link function;
Ordinary least squares.
基金项目: The research is supported by The Research Fund for the Doctoral Program of Higher Education of China
(20120092110021),National Natural Science Foundation of China (11171065),National Natural Science Foundation of China
(11201229),Natural Science Foundation of Jiangsu Province (BK2011058),Youth Foundation for Humanities and Social
Sciences Project from Ministry of Education of China (11YJC790311),and the Fundamental Research Funds for the Central
Universities.
作者简介: Huang Chao(1985-),male,PhD student,major research direction:statistics of extreme value. E-mail:
huangchao.******@. Correspondence author:Lin Jin-guan (1964-),male,professor,major research direction:
longitudinal data analysis, Copula. ******@seu..
-1-
0 Introduction
Copulas are functions that join multivariate distribu