I want to apply feature selection on a dataset with some 30-40K columns and 100 rows ( total size: 400MB-800MB ). To decrease the time consumed for calculations involved (feature-feature), I want to divide data in some 4-5 parts and execute all parts parallely. Since data is not huge in size, I am avoiding to use Hadoop. What can be the option for this parallelization.. ( multi-threading , GPU or anything else) ??

  • $\begingroup$ What do you want to do? I'm sure Bioconductor has multithreading functions for like gene counting. What're you trying to do? $\endgroup$
    – SmallChess
    Sep 16, 2016 at 11:03
  • $\begingroup$ i want to calculate mutual information (MI) between all possible pair of features for feature selection. $\endgroup$
    – phoenix
    Sep 16, 2016 at 13:40


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.