python - Custom folds for cross-validation in scikit-learn -
i use gridsearchcv (with n_jobs > 1) particular classifier, have information folds 10-fold cross-validation source. there way input data divided folds instead of using folds created gridsearchcv.
thanks!
you can create custom cv iterator such leaveonelabelout or leaveplabelout own precomputed folds encoded array of integers , pass cv iterator cv
argument of cross_val_score
, gridsearchcv
utilities.
Comments
Post a Comment