时变系数自回归模型参数的贝叶斯估计  

Bayesian estimation of autoregressive models with time-varying coefficients

作　　者：陈云仙 高星月 王钰莹 何幼桦[1] 

机构地区：[1]上海大学理学院,上海200444

出　　处：《上海大学学报（自然科学版）》2017年第5期732-741,共10页Journal of Shanghai University:Natural Science Edition

基　　金：国家自然科学基金资助项目(11371242)

摘　　要：针对时变系数的自回归模型,假设其系数的先后状态具有一定相关性.在仅有一条样本函数的情况下,运用贝叶斯方法给出一阶模型参数估计的表达式.通过数值模拟展示了模型系数的取值和样本容量对估计效果的影响.实证分析表明,该估计方法所得的统计结果能够较好地揭示实际问题的内在规律性.This paper analyzes a time-varying autoregression model where coeffcients are correlated at different time. When only one sample path is chosen, the Bayesian method is used for estimation. Formulas of estimation of the first order model are presented. This paper also discusses how the estimation is affected by the coeffcient values and the length of samples. To conclude, based on an empirical evidence, it is shown that the statistical results are consistent with the actual data.

关 键 词：时变系数 自回归模型 贝叶斯估计 人均GDP 

分 类 号：O212[理学—概率论与数理统计]