sklearn.metrics.multilabel_confusion_matrix

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

Compute class-wise (default) or sample-wise multi-label confusion matrix

When calculating class-wise multi_confusion, the input y_true and y_pred shape is (n_samples, n_labels) (multi-class multi-label case) or (n_samples,) (multi-class single-label case), and the output multi_confusion shape is (n_labels, 2, 2)

The above n_labels is the number of categories.

When calculating sample-wise multi_confusion, the input y_true and y_pred shape is (n_samples, n_labels) (multi-class multi-label case) or (n_samples,) (multi-class single-label case), and the output multi_confusion shape is (n_samples, 2, 2)

example:
multi-class single-label
>>> 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]]])

multi-class single-label
>>> 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]]])

reference:

sklearn.metrics.confusion_matrix — scikit-learn 1.1.1 documentation