Can we still use LASSO even when we have a small number of variables?

  • 3
    $\begingroup$ Why do you want to do variable selection? What do you want to achieve with your model? $\endgroup$ Oct 17, 2019 at 11:07
  • $\begingroup$ I want to remove redundant predictors? $\endgroup$
    – user30609
    Oct 17, 2019 at 12:03
  • $\begingroup$ Is the principle of LASSO the same with 30 predictors and 300 predictors and 3 predictors and 3000 predictors? $\endgroup$
    – user30609
    Oct 17, 2019 at 12:04
  • 2
    $\begingroup$ With 3 variables it is easy to do a best subset selection. It is in theory much better than LASSO, unless you also need parameter shrinkage. $\endgroup$
    – runr
    Oct 17, 2019 at 12:42
  • $\begingroup$ @user30609 What I meant to ask is whether you want to make inference about the relationship between the covariates and the outcome, or whether you are interested in making predictions. $\endgroup$ Oct 17, 2019 at 15:35


Your Answer

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

Browse other questions tagged or ask your own question.