Fig. 6
From: On the utilization of deep and ensemble learning to detect milk adulteration
ROC (Receiver Operating Characteristic) curves for binary and multiclass versions of (a) Random Forest, (b) Gradient Boosting Machine and (c) Convolutional Neural Network and their AUC (Area Under the ROC Curve) score. Multiclass ROC was calculated using micro-average strategy, that sums up the individual true positives, false positives, and false negatives for all classes