I am using nested CV for model evaluation and my target variable has imbalanced classes.
With Sklearn, I am using GridSearchCV and cross_val_score to perform the nested cross validation. Each takes a "cv" argument which can be a number or a specific cross validation method. In my case, I am passing the Stratified KFold object.
My question is, is it best practice to only pass the Stratified KFold object to GridSearchCV, or to cross val score only, or to both? For what reasons?