Bayesian Analysis of Multivariate Time Series Model including unobserved data and its application to macro economic analysis

我研究了使用最近开发的方法称为“贝叶斯方法”的时间序列数据的计量经济学分析，并将该方法应用于宏观经济分析。我们考虑了一个马尔可夫转换模型、一个具有多个结构突变的模型和一个具有不可观测风险溢价的模型，发现这些模型更符合实际，更便于分析。

I investigated econometric analysis of time series data using a recently developed method called "Bayesian approach", and applied the approach to macro economic analysis. We consider a Markov switching model, a model with multiple structural breaks, and a model with unobserved risk premium, and found that these models are more realistic and useful to analyse.