AUC is defined as ROC The area under the curve and the coordinate axis is ROC The curve is y=x above, so AUC The value range is 0.5 and 1 between.

AUC It can be used as an indicator of the quality of the model when comparing different classification models. Its main significance is AUC The larger the value, the higher the accuracy of the classifier.

Take the following figure as an example, the curve 1 Better than the curve 2

The standard for judging the quality of a classifier (prediction model) from AUC:

• AUC = 1, perfect classifier;
• AUC = [0.85, 0.95], very good results;
• AUC = [0.7, 0.85], average effect;
• AUC = [0.5, 0.7], low effect;
• AUC = 0.5, consistent with random guessing;
• AUC < 0.5, which is worse than random guessing.

There are two ways to calculate AUC, namely the trapezoidal method and the ROC AUCH method. Both methods use approximation to find approximate values.