I am trying to create a semiparametric model for university (we were told it HAS to be semiparametric) and I have 11 response variables, some of them categorical and the rest continuous. In the simple glm case I could easily decide which variables to include with a stepAIC function in R, but now I not only have to decide which of them to include, but which should be part of the nonparametric component and which to the parametric one and I can not find any way to make the decission.

Is there a good way to approach this or to gain some intuition?

Thank you very much

  • $\begingroup$ Even in the GLM case, stepwise regression has problems. $\endgroup$
    – Dave
    Aug 17, 2021 at 16:10
  • $\begingroup$ If you have binary variables, standard additive glm is already semiparametric. You are totally nonparametric in the additive component of the binary variable. Check mate. (Warning: your professor may not see it this way). $\endgroup$
    – user327671
    Aug 18, 2021 at 15:20


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.