我的 ROC 曲线的尺寸不正确

发布于 2025-01-10 04:42:34 字数 635 浏览 0 评论 0原文

我对机器学习非常陌生，我的任务是创建 ROC 曲线。我收到错误 ValueError: 发现样本数量不一致的输入变量: [200, 400]。这是我的 ROC 曲线代码的图片。

import sklearn.metrics as metrics
from sklearn.metrics import roc_curve, auc
y_pred = basic_model.predict(X_test).ravel()
nn_fpr_keras, nn_tpr_keras, nn_thresholds_keras = roc_curve(y_test, y_pred)
auc_keras = auc(nn_fpr_keras, nn_tpr_keras)
plt.plot(nn_fpr_keras, nn_tpr_keras, marker='.', label='Neural Network (auc = %0.3f)' % auc_keras)

我实际上知道问题是什么，我正在使用“sparse_categorical_crossentropy”来损失。（我的顾问说我必须使用它）。基本上他们还告诉我，现在我所要做的就是将我的 X_test 从 2D 数组更改为 1D 数组。我尝试使用 .flatten() 但这没有帮助？关于我应该做什么或我可以学习的任何资源有什么想法吗？

原文

im really new to machine learning and was tasked with creating a ROC curve. I get the error ValueError: Found input variables with inconsistent numbers of samples: [200, 400].
Here is a picture of my code for the ROC curve.

import sklearn.metrics as metrics
from sklearn.metrics import roc_curve, auc
y_pred = basic_model.predict(X_test).ravel()
nn_fpr_keras, nn_tpr_keras, nn_thresholds_keras = roc_curve(y_test, y_pred)
auc_keras = auc(nn_fpr_keras, nn_tpr_keras)
plt.plot(nn_fpr_keras, nn_tpr_keras, marker='.', label='Neural Network (auc = %0.3f)' % auc_keras)

I actually know what the problem is I am using 'sparse_categorical_crossentropy' for loss.(My advisor said I had to use it). Basically They also told me that now all I have to do is to change my X_test go from a 2D array to a 1D array. I have tried to use .flatten() but that didn't help? Any ideas on what I should do or any resources that I could study?

分享到QQ

分享到微博