i have created some new features for my model. I found people use kde plot to find out the correlation between the created feature and the target variable, but I am not really sure how to find the correlation from kde.

Any help on how to interpret the correlation from kde plot will very helpful

up vote 0 down vote accepted

CoreyLevinson on kaggle answered my question. I am simply adding the explaination.

The kde shows the density of the feature for each value of the target. There are usually 2 colored humps representing the 2 values of TARGET. If the humps are well-separated and non-overlapping, then there is a correlation with the TARGET. If the humps are overlapping a lot, then that means the feature is not well-correlated with the TARGET, because the TARGET is equally as common on those values of the feature.

Your Answer

 
discard

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.