基于混合Copula和Logistic回归的极端事件研究

发布时间：2018-04-03 15:09

本文选题：极端事件　切入点：混合Copula　出处：《苏州大学》2016年硕士论文

【摘要】：极端事件是指很少发生,然而一旦发生却产生极大影响的事件,因此关于极端事件的研究具有重要的实际意义。与单一的Copula函数相比,混合Copula函数可以通过自由选择不同类型的Copula函数来建立相关结构,能更准确地刻画极端事件发生的复杂关系。利用Gumbel Copula、Clayton Copula和Frank Copula构建的混合Copula函数,既可以涵盖单一上尾或下尾存在的情形,还可以展现上尾和下尾同时存在的情形。本文首次利用基于混合Copula函数的极端事件Logistic回归模型,将极端事件的影响因素和关联事件同时加入到模型中,可以利用条件概率去估计极端事件发生的概率。该模型充分利用了Copula函数在尾部相关性和Logistic回归在处理分类变量上的优势,优化了极端事件的概率估计和拟合精度。
[Abstract]:Extreme events refer to events that occur rarely, but once they occur, they have a great impact. Therefore, the study of extreme events has important practical significance.Compared with a single Copula function, the hybrid Copula function can establish the correlation structure by freely selecting different types of Copula functions, and can describe the complex relationship of extreme events more accurately.The hybrid Copula function constructed by Gumbel Copula Clayton Copula and Frank Copula can not only cover the existence of single upper tail or lower tail, but also show the existence of upper tail and lower tail.In this paper, the Logistic regression model of extreme events based on mixed Copula function is used for the first time. The influence factors of extreme events and the associated events are added to the model at the same time. The conditional probability can be used to estimate the probability of extreme events.The model makes full use of the advantages of Copula function in tail correlation and Logistic regression in dealing with classification variables, and optimizes the probability estimation and fitting accuracy of extreme events.
【学位授予单位】：苏州大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：F224

【相似文献】