两类重要的分布类及其应用-计算数学专业论文.docxVIP

ii ii 摘 要 关于分布类的研究是概率论中一个永恒的话题. 本文以特征函数为主要工具, 给出了 在特征函数中引入参数的一般方法, 提出了三参数 Linnik 分布和广义斜正态-正态分布两 种重要的分布类并讨论了这些分布类的几何封闭性和渐近性等性质, 最后讨论了三参数 Linnik 分布和广义斜正态-正态分布在自回归模型中的应用. 具体内容如下: §1 简要介绍了随机变量特征函数的相关概念. §2 讨论了在特征函数中引入新参数的一般方法, 提出了三参数 Linnik 分布和广义斜 正态-正态分布, 并讨论了它们的几何封闭性等性质. §3 讨论了三参数 Linnik 分布在自回归模型中的应用, 研究了以三参数 Linnik 分布为 边际分布的自回归模型. §4 讨论了广义斜正态-正态分布在自回归模型中的应用, 研究了以广义斜正态-正态 分布为边际分布的自回归模型一些重要性质. §5 提出了一些需要进一步研究的问题. 关键词: 特征函数; 广义 Linnik 分布; 斜正态-正态分布; 斜对称分布; 几何复合; 自 回归模型; 平稳性 iii iii Abstract The study of distribution classes is an eternal topic. In this paper, a general way of generalizing distribution is given by the toll-characteristic function. Two important dis- tribution classes called three-parameter Linnik distribution and generalized skew-normal- normal distribution are derived and some properties of these distributions and applications in autoregressive model are discussed. The main results of this paper can be summarized as following The notation of characteristic function are introduced. A general way of generalizing distribution is derived, three-parameter Linnik distri- bution and generalized skew normal-normal distribution are obtained, and their properties are discussed. The application of three parameters Linnik distribution in autoregressive model is discussed, autoregressive model with the distribution as marginal distribution is discussed. The application of generalized skew normal-normal distribution in autoregressive model is discussed, and properties of autoregressive model with the distribution as mar- ginal distribution are obtained. Some problems which will be studied in the future are pointed. Keywords: characteristic function; generalized Linnik distribution; skew normal- normal distribution; skewed distributions; geometric compound; autoregressive model; stationary property i i 独创性声明 本人声明所呈交的论文是我个人在导师指导下进行的研究工作及取得的研究成果. 尽我所知, 除了文中特别加以标注和致谢的地方外, 论文中不包括其他人己经发表或撰写 过的研究成果, 也不包含为获得西北师范大学或其他教育机构的学位或证书而使用过的 材料. 与我一同工作的同志对本研究所做的任何贡献均己在论文中作了明确的说明