I have a sample dataset with about a million records with about 7 feature variables. On running logistic regression on this dataset, I got 5 features with a positive coefficient and 2 with a negative coefficient.

Now I need to extrapolate this data to about 50 times and add a couple of feature variables to the set. How can I ensure that after generating random data and adding more features, I achieve maximum features with positive coefficients?

Is there a way to identify patterns or relationships that I could use to define rules for replicating the data in the way that I want ?

If this sounds a bit vague, I'd be happy to have a detailed discussion on this if possible.

  • $\begingroup$ Yes can you add more information about the "extrapolation"? $\endgroup$ – Pierre Lafortune Sep 16 '16 at 21:40
  • $\begingroup$ By extrapolation I just meant replicating the data to a significant size keeping the features intact $\endgroup$ – hbabbar Sep 20 '16 at 16:33

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.