Fig. 4
From: Efficiency of different measures for defining the applicability domain of classification models
Ames data set employing classification random forests. Cumulative accuracy (CA) curves are shown for all confidence measures and the two novelty measures \(\cos_{\alpha }\) and \(\gamma_{Euc}\). As with ROC curves, the inferior performance of the novelty measures can easily be seen. CA curves allow reading out the overall accuracy obtained when only a portion of x% of the data is predicted and the rest is rejected