scikit-learn中的cross_val_score函数可以通过交叉验证评估分数,非常方便,但是使用过程中发现一个问题,就是在cross_val_score的文档中对scoring的参数并没有说明清楚。
其实有另一个文档进行了说明:3.3. Metrics and scoring: quantifying the quality of predictionsscikit-learn.org
Scoring -Function--Comment
Classification
‘accuracy’
‘balanced_accuracy’
‘average_precision’
‘neg_brier_score’
‘f1’
for binary targets
‘f1_micro’
micro-averaged
‘f1_macro’
macro-averaged
‘f1_weighted’
weighted average
‘f1_samples’
by multilabel sample
‘neg_log_loss’
requires predict_proba support
‘precision’ etc.
suffixes apply as with ‘f1’
‘recall’ etc.
suffixes apply as with ‘f1’
‘jaccard’ etc.
suffixes apply as with ‘f1’
‘roc_auc’
‘roc_auc_ovr’
‘roc_auc_ovo’
‘roc_auc_