model_results = list() model_names = list() for model_name in models: model = models[model_name] k_fold = KFold(n_splits=folds, random_state=seed) results = cross_val_score(model, X_train, y_train, cv=k_fold, scoring=metric) code example
Example: sklearn kfold
from sklearn.model_selection import GridSearchCV
from sklearn.model_selection import KFold
# Regressor
lrg = LinearRegression()
#Param Grid
param_grid=[{
'normalize':[True, False]
}]
# Grid Search with KFold, not shuffled in this example
experiment_gscv = GridSearchCV(lrg, param_grid, \
cv=KFold(n_splits=4, shuffle=False), \
scoring='neg_mean_squared_error')