Area under the ROC Curve (AUC)
the accuracy of the machine learning tool
Technical definition
A measure of a classification model's performance, calculated as the proportion of the true positive rate to the false positive rate at a given false positive cost.