gaotao0727 发表于 2011-10-21 16:52 
多谢,我想到了,应该是极大似然估计~~~
Currently in SAS,
the default algorithm for iterative maximum likelihood is Fisher Scoring. People think, logistic model should be in the category of (generalized) linear model rather than in nonlinear model. That is how we name it. But general linear model is linear with mean whereas logistic model is linear with link function of mean, i.e. logistic function.
JingJu