class sklearn.metrics
方法
1.分类问题的度量
1 metrics.accuracy_score 2 metrics.auc 3 metrics.f1_score 4 metrics.precision_score 5 metrics.recall_score 6 metrics.roc_auc_score 7 ......
2.回归问题的度量
3.概率分布函数的度量
4.检索问题的度量
5.其他
查询地址：
https://scikit-learn.org/stable/modules/classes.html#sklearn-metrics-metrics
基于sklearn上聚类可使用的评估指标：
1 metrics.adjusted_mutual_info_score(…[, …]) 2 metrics.adjusted_rand_score(labels_true, …) 3 metrics.calinski_harabasz_score(X, labels) 4 metrics.davies_bouldin_score(X, labels) 5 metrics.completeness_score(labels_true, …) 6 metrics.cluster.contingency_matrix(…[, …]) 7 metrics.fowlkes_mallows_score(labels_true, …) 8 metrics.homogeneity_completeness_v_measure(…) 9 metrics.homogeneity_score(labels_true, …) 10 metrics.mutual_info_score(labels_true, …) 11 metrics.normalized_mutual_info_score(…[, …]) 12 metrics.silhouette_score(X, labels[, …]) 13 metrics.silhouette_samples(X, labels[, metric]) 14 metrics.v_measure_score(labels_true, labels_pred)
#大部分的评估指标都需要labels_true, 一些不需要labels_true指标如下