You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I just realized from here https://scikit-learn.org/stable/modules/cross_validation.html#computing-cross-validated-metrics, that "When the cv argument is an integer, cross_val_score uses the KFold or StratifiedKFold strategies by default, the latter being used if the estimator derives from ClassifierMixin."
So we would maybe like _PairsClassifierMixin to inherit from ClassifierMixin to have a nice stratification based on the class when doing cross-validation. We could also maybe add an explanatory paragraph about that in the docs.
Note however that ClassifierMixin implements a default score method that is accuracy. I guess it's OK to overwrite it with our own default score (roc_auc) in the end, but it's good to clarify that.
The text was updated successfully, but these errors were encountered:
I just realized from here https://scikit-learn.org/stable/modules/cross_validation.html#computing-cross-validated-metrics, that "When the cv argument is an integer, cross_val_score uses the KFold or StratifiedKFold strategies by default, the latter being used if the estimator derives from ClassifierMixin."
So we would maybe like
_PairsClassifierMixin
to inherit fromClassifierMixin
to have a nice stratification based on the class when doing cross-validation. We could also maybe add an explanatory paragraph about that in the docs.Note however that
ClassifierMixin
implements a default score method that is accuracy. I guess it's OK to overwrite it with our own default score (roc_auc) in the end, but it's good to clarify that.The text was updated successfully, but these errors were encountered: