基于吉伯斯样本生成器的向量自回归模型选择

 内容提要：向量自回归模型是多元时间序列分析中最常用的方法之一。在建立模型的过程中模型选择是非常重要的一个环节，如果候选模型不是很多时，可以通过比较每个模型的准则值如AIC、AICc、BIC或HQ进行模型选择。可是，当存在大量候选模型时，我们无法一一比较每个模型的准则值。为了解决这个问题，本文提出一个基于吉伯斯样本生成器的向量自回归模型选择方法，结果表明应用该方法能够从大量候选模型中准确、高效地确认准则值最小的模型。

Vector Autoregressive Model Selection Based on Gibbs Sampler

Model selection is an important step in VAR modeling,in the process of Model selection,when the number of candidate models is not too large,we can select model by comparing the criterion value of each model,such as AIC,AICc,BIC or HQ.However,when there are a large number of candidate models,we can't compare the criterion value one by one,so we put forward a model selection procedure based on Gibbs sampler to solve this problem,this method can accurately and efficiently identify VAR model with the smallest criterion value.