I want to visualize a data set for EDA having 8 parameters and 2 class labels. I am confused which parameters to select and on what basis? Also is I want to know whether it would be better to apply dimensionality reduction techniques before visualization?


Faraz! Since you have just 8 parameters it is not time consuming to observe all you parameters and how they are represented for both classes.

You can start with simple information about distribution of parameters over your 2 classes. Check this notebook where an author has explored Titanic data. You can start your explanation in the similar way.

For more tips from a community you need to provide us with more data like what is the size of your data set, are the parameters categorical or numerical, etc. And actually what is the purpose of your EDA? What is your next step going to be?

  • $\begingroup$ It means for me you have overlapping classes. Let say op1 = Sport clothes, op2 = Casual clothes but there are some goods that can be in both of categories. So you can now visualize how often the clothes is from cotton/polyester/etc depending on the class. And so on.... $\endgroup$ – zina Jul 2 '19 at 14:50

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.