Skip to main content

Table 1 Classification performance of each model (mean \(\pm\) std)

From: GAT-LI: a graph attention network based learning and interpreting method for functional brain network classification

Model Accuracy Sensitivity Specificity F1 AUC MCC
SVM 0.6618 ± 0.0110 0.6515 ± 0.0413 0.6717 ± 0.0218 0.6521 ± 0.0211 0.7170 ± 0.0188 0.3238 ± 0.0230
PCA + SVM 0.6686 ± 0.0195 0.6554 ± 0.0561 0.6811 ± 0.0334 0.6576 ± 0.0300 0.7184 ± 0.0156 0.2793 ± 0.0339
RF 0.6599 ± 0.0309 0.5921 ± 0.0309 0.7245 ± 0.0324 0.6295 ± 0.0330 0.7153 ± 0.0325 0.2978 ± 0.0768
MLP 0.6754 ± 0.0309 0.6634 ± 0.0401 0.6868 ± 0.0601 0.6660 ± 0.0297 0.7535 ± 0.0297 0.2899 ± 0.0612
CNN 0.6550 ± 0.0312 0.6316 ± 0.0466 0.6774 ± 0.0345 0.6407 ± 0.0364 0.7111 ± 0.0314 0.3098 ± 0.0615
GCN-at (1st-order) 0.5971 ± 0.0460 0.6059 ± 0.0398 0.5887 ± 0.0619 0.5951 ± 0.0417 0.6537 ± 0.0503 0.2775 ± 0.0645
GCN-at (Cheby) 0.6357 ± 0.0217 0.6812 ± 0.0558 0.5925 ± 0.0558 0.6452 ± 0.0262 0.6926 ± 0.0368 0.2975 ± 0.0600
GAT-fc 0.6184 ± 0.0332 0.7089 ± 0.0507 0.5321 ± 0.0927 0.6445 ± 0.0209 0.6547 ± 0.0426 0.3155 ± 0.0713
GAT-average 0.6734 ± 0.0354 0.7386 ± 0.0270 0.6113 ± 0.0801 0.6889 ± 0.0226 0.7361 ± 0.0321 0.3237 ± 0.0621
GAT-learn 0.5845 ± 0.0371 0.6000 ± 0.1765 0.5698 ± 0.1473 0.5732 ± 0.0844 0.5849 ± 0.0385 0.1798 ± 0.0821
GAT2 0.6802 ± 0.0269 0.7406 ± 0.0408 0.6226 ± 0.0534 0.6931 ± 0.0248 0.7358 ± 0.0373 0.3426 ± 0.0628
  1. The bold means it is the best result for each metric (column of the table)