基于Bayesian思想的最优Copula函数选择  

Optimal Copula Function Selection Based on Bayesian Methodology

作　　者：樊妮[1] 赫孝良[1] 赵谦[1] 

机构地区：[1]西安交通大学数学与统计学院,西安710049

出　　处：《工程数学学报》2012年第4期516-522,共7页Chinese Journal of Engineering Mathematics

摘　　要：作为一种有效的数据建模工具,Copula在金融领域中得到了广泛应用.本文针对如何选择最优的Copula函数这一在应用中亟待解决的问题进行研究,基于Bayesian思想提出了一种选择方法.所提方法具有良好的理论基础,且应用方便.在模拟数据和原油市场数据上的实验表明,所提方法具有良好的最优Copula函数选择效果,明显优于基于K-S检验的方法.As an efficient tool for data modeling, Copula has been widely applied in financial analysis. In this paper, a new method based on Bayesian statistics is proposed to deal with the optimal Copula selection problem. The proposed method is both of good theoretical properties and convenient to be implemented. Numerical experiments on both simulated and real crude oil market data demonstrate that the proposed method is promising for optimal Copula selection and notably effective than the K-S test method.

关 键 词：COPULAS Bayes定理 选择准则 K—S检验 

分 类 号：O212.8[理学—概率论与数理统计]