Fig. 1.

ROC curves generated from training and testing the classifier based on the full dataset (black), 10-fold cross-validation (green) and 3-fold cross-validation (red). Note the modest differences between testing based on the full dataset and cross-validation. AUC is shown in Table 2.