0
$\begingroup$

I am estimating a conditional ridge logit model, there is very little bibliography about it, and I do not know if the constant coefficients are penalized. My model has 2 variables and 3 alternatives, so I want to estimate 8 coefficients of which there are 2 constant coefficients, 6 coefficients of the 2 variables of the 3 modes.

If you could help me with bibliography better yet.

Thank you!

$\endgroup$
0
$\begingroup$

Look at the ESL book The Elements of Statistical Learning II, page 61, formula (3.41). The constant is not penalized.

$\endgroup$
  • 1
    $\begingroup$ Please post a snapshot or LaTeX transcription of the formula so the answer remains useful if the link breaks/in the absence of the book ... $\endgroup$ – Ben Bolker Mar 31 at 0:52
  • 1
    $\begingroup$ but in that book it refers to linear regression, does this also apply to the conditional ridge logit model? $\endgroup$ – lampard811 Mar 31 at 0:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.