Please suggest a feature selection technique which selects features such that they explain target well but are not correlated with given know confounds.

This paper(section 2.3) suggests using a GAN for the same. I would prefer something other than a deep learning technique.


Not sure if you are looking for hyper parameter tuning, can you describe your dataset and the number of features it has.

An alternate approach would be to use feature extraction from sklearn to build sparse matrices rather than keeping your categorical variables as they are, it would give you can idea which features are important.

A tutorial can be found here :


| improve this answer | |
  • $\begingroup$ this doesn't answer my question $\endgroup$ – claudius Mar 22 '18 at 4:21

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.