zhourongwang 发表于 2010-10-16 20:51 
在SAS做完Logistic回归后，结果中的AIC与SC越小越好，但是具体小到什么程度才算好呢？有没有这方面的资料啊？请教各位大侠！！！
In the general, the AIC is 2k-2log(L)
where
k is the number of parameters and L
is the evaluated maximized value of the likelihood function for the estimated model.
It depents on k, n sample size and functional form of likelihood function. With a given sample size and model function is correct specified, you can use it to evaluate which one is better.
For example model (A) using variables of x1 x2 x3 is better than a model (B) using variables of x1 x2 x3 x4.
扫码加好友，拉您进群
全部版块
我的主页
收藏
