I'm just doing some interview preparation for a data science interview and this question came up.

I'm familiar with general feature selection methods such as best subset selection, forward stepwise, backward stepwise, and lasso regression to help select most important feature. However, these methods work in general, and I'm not sure if they work particularly well with sparse matrices and/or if there is something better.

Are there other considerations in particular for sparse matrices?

Thank you in advance for your help!

  • 1
    $\begingroup$ if you're interested in feature selection, your title should be reworded because control the number of features is vague $\endgroup$ – Antoine Aug 7 '15 at 7:08
  • $\begingroup$ Lasso works well for sparse matrices, glmnet has been optimized to handle sparseness well. In my opinion, this is an awful interview question, as it is asking for fact regurgitation. You are more likely to encounter an interviewer setting up an imaginary situation and then talking through it with you. What method is good to use is very sensitive to the particulars and constraints of the problem. $\endgroup$ – Matthew Drury Aug 7 '15 at 15:21

Your Answer

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

Browse other questions tagged or ask your own question.