Fig. 2
From: LightCUD: a program for diagnosing IBD based on human gut microbiome data
Evaluation of the performance of LightCUD. We evaluated the accuracy of disease classification using LightCUD with receiver operating characteristic curve and precision-recall curves representing the results. Lines in each subplot with different colors represent the model performance in one of the five-fold cross validations. As the training sets were unbalanced, we reported both the AUC values and the AP values. The average AUC and AP were labeled under corresponding curves