sklearn - 对类的子集进行精确评分的交叉验证

Question

我有一个分类数据集,有3个类标签[0,1,2].

我想运行交叉验证并尝试几个估算器,但我对只有1级和2级精度的评分感兴趣.我不关心0级的精度,我不希望它的得分甩开CV优化.我也不关心任何课程的召回.换句话说,我想确保无论何时预测1或2,它都具有很高的置信度.

所以问题是,如何运行cross_val_score并告诉其评分函数忽略0级的精度？

更新:根据接受的答案,这是一个示例答案代码:

def custom_precision_score(y_true,y_pred):
  precision_tuple, recall_tuple, fscore_tuple, support_tuple = metrics.precision_recall_fscore_support(y_true, y_pred)
  precision_tuple = precision_tuple[1:]
  support_tuple = support_tuple[1:]
  weighted_precision = np.average(precision_tuple, weights=support_tuple)
  return weighted_precision

custom_scorer = metrics.make_scorer(custom_precision_score)

scores = cross_validation.cross_val_score(clf, featuresArray, targetArray, cv=10, scoring=custom_scorer)

Answer 1

cross_val_score包括射手调用对象，可以用自己的测试策略设置使用make_scorer。并且您可以在自定义评分函数中设置您要测试的组，该函数score_func(y, y_pred, **kwargs)由make_scorer.