分类问题:
‘accuracy’----------------------------metrics.accuracy_score
‘average_precision’---------------metrics.average_precision_score
‘f1’-------------------------------------metrics.f1_score
‘f1_micro’----------------------------metrics.f1_score
‘f1_macro’---------------------------metrics.f1_score
‘f1_weighted’-----------------------metrics.f1_score
‘f1_samples’------------------------metrics.f1_score
‘neg_log_loss’----------------------metrics.log_loss
‘precision’ ---------------------------metrics.precision_score
‘recall’ --------------------------------metrics.recall_score
‘roc_auc’-----------------------------metrics.roc_auc_score
回归问题
‘neg_mean_absolute_error’----metrics.mean_absolute_error
‘neg_mean_squared_error’-----metrics.mean_squared_error
‘neg_median_absolute_error’–metrics.median_absolute_error
‘r2’-------------------------------------metrics.r2_score
sklearn.metrics官方介绍网址
举例:
parameters = {'kernel':['linear','poly','rbf'], 'gamma':np.logspace(-5, 20, num=5, base=2.0),'C':np.logspace(1, 6, num=6, base=2.0)}
grid_search = GridSearchCV(svm.SVR(), parameters, cv=10, n_jobs=10, scoring='f1')