Figure 3
From: The influence of negative training set size on machine learning-based virtual screening
The changes of the ROC curves and AUC values for different number of inactives used for ML training. The figure presents ROC curves for 5-HT1A target illustrating the improvement of the machine learning models performance, calculated for several negative training example quantities (i.e. 100, 1000, 2000, 3000 and 4000) (A) and the dependence of AUC on the number of inactive training examples (B).