Correlation Handling #288
Correlation Handling
#288
-
|
Does the library offer any methods or options to analyze or address variable correlation both pre and post binning? I want to drop highly correlated variables. |
Beta Was this translation helpful? Give feedback.
Answered by
guillermo-navas-palencia
Dec 6, 2023
Replies: 1 comment
-
|
Hi @maxsanwal. The OptBinning library does not offer such methods. For analyzing correlation post-binning, you could employ rank correlation methods or simply Pearson's correlation. These methods also apply to pre-binning numerical variables. However, for categorical variables, other methods are required. |
Beta Was this translation helpful? Give feedback.
0 replies
Answer selected by
maxsanwal
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Hi @maxsanwal.
The OptBinning library does not offer such methods. For analyzing correlation post-binning, you could employ rank correlation methods or simply Pearson's correlation. These methods also apply to pre-binning numerical variables. However, for categorical variables, other methods are required.