I am trying to fit a multinomial logistic regression model. I have four outcomes. I also need to do a variable selection. I have about 15 variables. I know that in the model statement, if I use link=glogit
then it will perform multinomial logistic regression. However, is there a way to do variable selection also?
0
votes
1 Answers
0
votes
proc logistic
has a few different variable selection methods that can be specified in the model
statement. See Table 60.8 Effect Selection Options in the documentation. Many of these options and syntax are shared with other procedures, such as proc glmselect
and proc reg
.
They provide a Stepwise Selection example that shows one way to use automatic variable selection.
proc logistic data=Remission outest=betas covout;
model remiss(event='1')=cell smear infil li blast temp
/ selection=stepwise
slentry=0.3
slstay=0.35
details
lackfit;
output out=pred p=phat lower=lcl upper=ucl
predprob=(individual crossvalidate);
run;