0
$\begingroup$

I'm new and have searched many questions about this problem in this stack, but those answers aren't clear enough for me.

The point is the area under PR curve of my binary classes is the same as the version I took SMOTE.

Here is my pseudo code:

X_train, y_train, X_test, y_test = split(data)
smote_X_train, smote_y_train = SMOTE(X_train, y_train)
myclassifier.fit(smote_X_train, smote_y_train)
myclassifire.predit(X_test)

And here is the result that I've got:

Yellow - SMOTE

Blue - Imbalanced classes

Should yellow increase, right? This confuses me.

I guess the possible causes of this is either an algorithm or features. Am I right?

$\endgroup$
0
$\begingroup$

It looks like you have evidence that you did not need to balance your classes.

Class balancing is a generally overused technique, you shouldn't expect it to always help, and should avoid it until you've shown its necessary.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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