I have a large dataset with 800 columns and 6,000,000 rows with many dummy variables (70%+). I want to Normalize it. Given that so many variables are binary, taking values 0 or 1, I am tending towards MinMax Normalization which will render all continuous variables in the [0, 1] range and leave the dummy variables intact. On the contrary Normalized continuous variables will range mostly in the interval [-2, 2] and there is an issue with the dummy variables, whether they should be affected.

So, to conclude, I would appreciate views on the Pros and Cons of the Two Normalization methods --MinMax and Standardization-- in the Setting I described. I say views because perhaps there is no one single correct answer.

  • $\begingroup$ Why are you asking? What are you going to do with your data so that you think to make binary and continuous variables "comparable" in scale? $\endgroup$ – ttnphns Aug 23 '17 at 20:45

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.