Yes, it compares to all the other variabels together
For example,a categorical variable maritalstatus (single,married,divorced) coded as 0,1,2 respectivelly and lasso chose 2.maritalstatus then you say the odds of the outcome... while being divorced is (e^coefficient) times the outcome when a person is single or married
↧
Re: An introduction to the lasso in Stata
↧
