Using mutual_info_classif for Continuous Features and Discrete Target #27284
Unanswered
RafiullahOmar
asked this question in
Q&A
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
I came across this statement in Google Bard, but I couldn't find any reference to it in the scikit-learn (sklearn) documentation. The statement is as follows: 'The mutual information between a continuous feature and a discrete target variable is not always meaningful. The mutual information measures the amount of information that one variable can provide about another variable. In the case of a continuous feature and a discrete target variable, the continuous feature cannot provide any information about the discrete target variable, because the discrete target variable can only take on a limited number of values.'
I am interested in using the mutual_info_classif scoring function for continuous features and a discrete target. It seems to work, but I received a warning that states: 'UserWarning: Clustering metrics expect discrete values but received continuous values for label, and binary values for target'.
Beta Was this translation helpful? Give feedback.
All reactions