Skip to main content
. 2020 Sep 29;14:577887. doi: 10.3389/fnins.2020.577887

TABLE 6.

The evaluation of classification performance corresponding to different functional connectome features.

Methods Accuracy (%) Sensitivity (%) Specificity (%) AUC
Connection (C) 64.29 50.00 77.23 0.728
Global Metrics (G) 69.05 65.00 72.73 0.793
Nodal Metrics (N) 73.81 80.00 68.18 0.865
MK_C + G 71.43 75.00 68.18 0.836
MK_C + N 78.57 85.00 72.73 0.888
MK_G + N 80.95 85.00 77.27 0.918
MK_C + G + N 83.33 90.00 77.27 0.927

MK-SVM, multiple kernel-support vector machine.