Using GridsearchCV () with holdout validation

Question

GridsearchCV () has an argument cv whose value by default is 3 means that it is 3fold. Is there any way to use Gridsearch() with a holdout validation scheme. For example 80-20% split???

Answer 1

You could use sklearn.model_selection.ShuffleSplit or sklearn.model_selection.StratifiedShuffleSplit with n_splits=1 in place of cv argument

Read the documentation for more details http://scikit-learn.org/stable/modules/generated/sklearn.model_selection.StratifiedShuffleSplit.html#sklearn.model_selection.StratifiedShuffleSplit