Copula一词原意为连接，它把多个随机变量的边缘分布连接在一起形成联合分布。变量间的相关结构完全由Copula决定，而各变量的统计特征由其边缘分布确定。与我们描述变量问相关关系常用的相关性相比，Copula描述的多元随机变量间的相关结构可以提供更准确的信息，目前Copula已经成为流行的多变量建模工具。(Copula is originally meant to connect the edge distributions of multiple random variables together to form a joint distribution. The correlation structure between variables is completely determined by Copula, and the statistical characteristics of each variable are determined by its edge distribution. Copula is a popular tool for multivariate modeling, which can provide more accurate information than the commonly used correlations for describing intervariate relationships.)