随时间变化系数参数AR模型的Bayes估计

摘要

采用Bayes方法,给出一维随时间变化系数自回归时间序列模型(TVPAR模型)中各参数的Bayes估计.选取未知参数的先验分布为均匀分布和逆伽马分布,采用MCEM算法,给出了模型中各参数的估计算法,并对模型进行了模拟计算与实证分析.%The authors proposed the Bayesian method to estimate the parameters in a special time vary parameter autoregression model ( TVPAR model). Assuming that coefficient parameters are uniform, and variance is inverse gamma distribution, we got the iterate steps of all parameters by means of MCEM algorithm. Simulation and a real data analysis were performed to illustrate the proposed estimation method.