文档介绍：摘要近年来,时间序列分析方法的研究和应用飞速发展,特别在经济领域,,,由于经济领域的特殊性,,. 本文主要研究了向量自回归移动平均模型(VARMA)和向量分整自回归移动平均模型(VARFIMA)的贝叶斯推断理论及其应用. 首先,进行了时间序列VARMA模型的贝叶斯分析,分析了时间序列 VARMA(p,g)模型的统计结构及其条件似然函数,,, 并利用WinBUGS进行VARMA模型仿真分析. 其次, VARFIMA(p,d,g)模型的统计结构开始,构建了模型的似然函数和参数的先验分布,严密地推导了模型参数的条件后验分布密度函数;利用一组用MATLAB软件模拟的二维长记忆时间序列,通过WinBUGS进行仿真分析. 关键词:时间序列;贝叶斯推断;MCMC方法;G bbs抽样;Wi nBUGS Abs仃act Recent years have witnessed wide applications oftime series analysis inmany inthe economic field,more and more practitioners have given intensive research totime seriesanalysismethods tomake full use the deepening ofreform and therapid development ofeconomy,there is alarge need for dataanalysis andprocessing intheeconomic field ,in practical applications,due totheparticularity oftheeconomic field,the use offrequency often encounter many difficulties in traditional statisticalmethods foreconomic time series model ,this thesisintroduces anew model foreconomic timeseries analysis of analysis methods provide amore rational analytical framework foreconomictime seriesmodels. TheBayesian inference theory andapplications ofthe vector autoregressive moving average model fVARMA)and thewhole sub-vector autoregressive moving average model(VARFIMA)are mainly studied. Firstly,atime model、析tllBayesian methods the statistical structure ofthemodel and itslikelihood function areanalyzed,according to thelikelihood function theprior distribution ofmodelparameters normal—Gamma prior distribution ofBayesian inference thebasis ofthe theory derived fromstatisticalmethods thedistribution oftheforecast ispredicted.