Fig. 4.

ROC-AUC of our model compared with those reported by previous studies. (Left) ROC-AUC averaged for 12 subdatasets were compared between our model (blue) and previous studies (gray). (Right) ROC-AUC of our model for each subdataset