Skip to main content
. 2020 Jul 6;8:e9470. doi: 10.7717/peerj.9470

Table 5. The testing results of CNN_3 and CNN_4 performed on the combined dataset.

The testing results performed on the hold-out data.

Site Ratio Model
CNN_3 CNN_4
AUPRC Precision Recall MCC AUPRC Precision Recall MCC
Donor 1:1 0.801 0.394 0.963 0.600 0.802 0.387 0.964 0.595
1:3 0.847 0.619 0.917 0.741 0.848 0.566 0.940 0.716
1:5 0.887 0.758 0.885 0.807 0.887 0.705 0.921 0.792
1:7 0.934 0.888 0.850 0.854 0.934 0.852 0.901 0.860
Acceptor 1:1 0.674 0.277 0.933 0.487 0.679 0.266 0.941 0.475
1:3 0.734 0.475 0.871 0.623 0.738 0.463 0.881 0.618
1:5 0.801 0.675 0.804 0.717 0.802 0.631 0.845 0.709
1:7 0.886 0.850 0.761 0.776 0.888 0.828 0.805 0.788

Notes.

Bold styling emphasizes the highest values regarding the evaluation metrics used in the study.