Hi Khuyen Tran,

Ok so I checked the documentation of YellowBrick. And I think there is a misunderstanding. Rank1D does not assess the feature importance relative to the target. As quoted from the documentation :

"A one-dimensional ranking of features utilizes a ranking algorithm that takes into account only a single feature at a time (e.g. histogram analysis). By default we utilize the Shapiro-Wilk algorithm to assess the normality of the distribution of instances with respect to the feature."

It only assess the normality of the features with respect to the target. So I guess this is usefull in a scenario when you want to assess the validity of the hypotheses of your linear regression model. But having a feature normaly distributed with respect to the target is not sufficient (at all) to make it a good predictor :-)

I would recommend using Rank2D that computes the pearson correlation between pair of features and check the correlation with the target instead.

Cheers

I am a machine learning engineer at Armis and I love to learn and share my passion for data science — https://www.linkedin.com/in/adrien-biarnes-81975717

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store