zkarl 发表于 2011-11-22 10:48 
谢谢啊~能还能具体解释下吗,很感激
Likelihood ratio: 一种类似于F 检验的统计量,不过这个检验是用于多个等式的
Denote the variance-covariance matrix of residuals asΣ, the LR test is given by:
LR=T[log|Σr| −log|Σu|]
The LR statistics is asymptotically distributed as a chi-square distribution with degrees of freedom equal to the total number of restrictions.
Information criteria:
MAIC = log|Σ|+ 2k'/T
MSBIC = log|Σ|+k′log(T)/T
MHQIC = log|Σ|+2k′log(log(T)) /T
另:其实这两种方法都是用来选择最优滞后阶数的,当选择了最优的滞后阶数,实际上就是对于不显著的滞后项的剔除。
如果你一定要“剔除”的话,可能LR更符合你心中的“剔除”的概念。