首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于正态-混合对数正态共轭先验下时序趋势项的贝叶斯推断
引用本文:李洪毅,欧祖军,石艳平.基于正态-混合对数正态共轭先验下时序趋势项的贝叶斯推断[J].甘肃联合大学学报(自然科学版),2007,21(4):36-39.
作者姓名:李洪毅  欧祖军  石艳平
作者单位:1. 吉首大学,师范学院,湖南,吉首,416000
2. 吉首大学,数计学院,湖南,吉首,416000
基金项目:吉首大学校科研和教改项目
摘    要:系统地分析了AR(p)误差项的时间序列模型及条件似然函数,并根据似然函数的统计结构构造了模型参数的共轭先验分布,研究了正态-混合对数正态共轭先验下模型的贝叶斯推断理论,包括趋势项的核估计参数及先验参数的后验分布的统计推断.

关 键 词:AR(p)误差  Gibbs抽样  贝叶斯估计
文章编号:1672-691X(2007)04-0036-04
修稿时间:2007-03-30

Bayesian Estimation of Time Series Model with Trend Based on Conjugate prior of Normal-Mixed Logarithmic Normal Distribution
LI Hong-yi,OU Zu-jun,SHI Yan-ping.Bayesian Estimation of Time Series Model with Trend Based on Conjugate prior of Normal-Mixed Logarithmic Normal Distribution[J].Journal of Gansu Lianhe University :Natural Sciences,2007,21(4):36-39.
Authors:LI Hong-yi  OU Zu-jun  SHI Yan-ping
Institution:1. Normal School of Jishou University,Jishou 416000 ,China; 2. School of Mathematics and Computer Seienee,Jishou University,Jishou 416000,China
Abstract:This paper analyzes the mathematic model and conditional likelihood of time series model with AR(p) error,and constructs conjugate prior distribution of parameters.Based on the conjugate prior of normal-mixed logarithmic normal distribution,the authors study the Bayesian inference of the model,including the statistical inference of the parameter of the kernel estimation of trend and the prior parameters of posterior distribution.
Keywords:AR(p) error  Gibbs sampling  Bayesian estimation
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号