feature-selector icon indicating copy to clipboard operation
feature-selector copied to clipboard

can you clarify how you remove correlated features

Open Sandy4321 opened this issue 4 years ago • 3 comments

can you clarify how you remove correlated features

as written in https://towardsdatascience.com/a-feature-selection-tool-for-machine-learning-in-python-b64dd23710f0

For each pair of correlated features, it identifies one of the features for removal (since we only need to remove one

so you just remove one from pair ?

Sandy4321 avatar Jan 04 '21 15:01 Sandy4321

Most variables are correlated with each other and thus they are highly redundant, let's say if you have two variables that are highly correlated, keeping the only one will help in dimensionality reduction and it doesn't cause that much loss of information.

One Question may arise you, Which Variable to keep? Keep the one that has a higher correlation with the target variable.

ayush714 avatar Apr 09 '21 03:04 ayush714

I see but Collinear Features how you calculated collinearity for categorical values ?

Sandy4321 avatar Apr 09 '21 14:04 Sandy4321

Hi Sandy4321, I found a brilliant article that will help with your question : https://towardsdatascience.com/the-search-for-categorical-correlation-a1cf7f1888c9

rajlm10 avatar May 20 '21 09:05 rajlm10