获取等于 class 标签数量的多 class 混淆矩阵

Question

我在 sklearn 中训练了随机森林分类器来预测多 class class 化问题。

我的数据集有四个 class 标签。但是我的代码创建了 2x2 混淆矩阵

y_predict = rf.predict(X_test)
conf_mat = sklearn.metrics.confusion_matrix(y_test, y_predict)
print(conf_mat)

输出：

[[0,   0]

 [394, 39]]

如何获得 4x4 混淆矩阵来分析 TP、TN、FP、FN。

Answer 1

来自
的文档 http://scikit-learn.org/stable/modules/generated/sklearn.metrics.confusion_matrix.html

y_true = ["cat", "ant", "cat", "cat", "ant", "bird"]
y_pred = ["ant", "ant", "cat", "cat", "ant", "cat"]
confusion_matrix(y_true, y_pred, labels=["ant", "bird", "cat"])

结果：

array([[2, 0, 0], 
       [0, 0, 1], 
       [1, 0, 2]])

获取等于 class 标签数量的多 class 混淆矩阵

Get multi-class confusion matrix equal to number of class labels

classification

confusion-matrix

scikit-learn