[英]scikit-learn - explained_variance_score
I'm using scikit-learn to build a sample classifier which was trained and tested by an svm.我正在使用 scikit-learn 构建一个样本分类器,该分类器由 svm 训练和测试。 Now i want to analyze the classifier and found the explained_variance_score but i don't understand this score.
现在我想分析分类器并找到了explained_variance_score,但我不明白这个分数。 For eg I get the classification report of the clf and it looks like this...
例如,我得到了 clf 的分类报告,它看起来像这样......
precision recall f1-score support
0.0 0.80 0.80 0.80 10
1.0 0.80 0.80 0.80 10
avg / total 0.80 0.80 0.80 20
not bad but the EVS is only 0.2
...sometimes its -0.X
...so how could this happen?不错,但 EVS 仅为
0.2
...有时是-0.X
...那么这怎么会发生呢? Is it important to have an good EVS?拥有一个好的 EVS 很重要吗? maybe someone could explain me this...
也许有人可以向我解释这个...
Y_true and Y_pred: Y_true 和 Y_pred:
[ 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 0. 0. 0. 0. 0. 0. 0. 0.
0. 0.]
[ 1. 1. 1. 1. 1. 0. 0. 1. 1. 1. 1. 0. 0. 0. 0. 0. 1. 0.
0. 0.]
Explained variance is a regression metric, this not well defined for the classification problem, there is no point in applying this for such testing.解释方差是一个回归度量,这对于分类问题没有很好地定义,将其应用于此类测试是没有意义的。 This is a method for validating models like Support Vector Regression, Linear Regression, etc.
这是一种验证支持向量回归、线性回归等模型的方法。
explained_variance_score, EVS tells you how much variance is explained by your model. Explain_variance_score,EVS 告诉您模型解释了多少方差。 The maximum value is one.
最大值为一。 Higher the EVS better is your model.
EVS 越高越好是您的模型。
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.