Hands-On Artificial Intelligence for Cybersecurity

ROC curve

Often, between false positives and false negatives, there is a trade-off; reducing the number of false negatives leads to an increase in false positives and to detect the existence of this trade-off, a particular curve is used, known as the ROC curve. This is as shown in the following image:

(Image Credit: Wikipedia, at https://commons.wikimedia.org/wiki/File:ROC_curve.svg )

The ROC curve is calculated using roc_curve() of scikit-learn, which takes the target values and the corresponding probabilities as parameters as shown in the following code:

from sklearn.metrics import roc_curveFPR, TPR, OPC = roc_curve(targets, probs)