sklearn.metrics.multilabel_confusion_matrix

sklearn.metrics.multilabel_confusion_matrix(y_true, y_pred, *, sample_weight=None, labels=None, samplewise=False)

计算class-wise（默认）或sample-wise多标签混淆矩阵

计算class-wise multi_confusion时，输入y_true和y_pred形状为(n_samples, n_labels) （多类多标签情况）or (n_samples,)（多类单标签情况），输出multi_confusion形状为(n_labels, 2, 2)

上述n_labels即类别数。

计算sample-wise multi_confusion时，输入y_true和y_pred形状为(n_samples, n_labels) （多类多标签情况）or (n_samples,)（多类单标签情况），输出multi_confusion形状为(n_samples, 2, 2)

例子：
多类单标签
>>> import numpy as np
>>> from sklearn.metrics import multilabel_confusion_matrix
>>> y_true = np.array([[1, 0, 1],
...                    [0, 1, 0]])
>>> y_pred = np.array([[1, 0, 0],
...                    [0, 1, 1]])
>>> multilabel_confusion_matrix(y_true, y_pred)
array([[[1, 0],
        [0, 1]],

       [[1, 0],
        [0, 1]],

       [[0, 1],
        [1, 0]]])

多类单标签
>>> y_true = ["cat", "ant", "cat", "cat", "ant", "bird"]
>>> y_pred = ["ant", "ant", "cat", "cat", "ant", "cat"]
>>> multilabel_confusion_matrix(y_true, y_pred,
...                             labels=["ant", "bird", "cat"])
array([[[3, 1],
        [0, 2]],

       [[5, 0],
        [1, 0]],

       [[2, 1],
        [1, 2]]])

参考：

sklearn.metrics.confusion_matrix — scikit-learn 1.1.1 documentation