Figure 6.

Association between the sample sizes and the area-under-the-curve values of the classifier models (A). The receiver operating characteristic curve of all of the models constructed based on the original datasets (B).