Skip to main content
. 2020 Aug 17;15(8):e0237587. doi: 10.1371/journal.pone.0237587

Fig 5. Graphic user interface of Result Inspection part of FAE.

Fig 5

Common clinical statistics for each model are shown on training dataset and validation dataset. Models are compared using ROC curve, AUC of each model, and feature contributions. FAE can also create high-resolution plots and generate brief description of algorithms used in selected model.