Skip to main content
. 2022 Feb 4;15:828512. doi: 10.3389/fnins.2021.828512

Table 2.

Comparisons of different methods.

Model Accuracy AUC Precision Recall F1-score
SVM 66.0 ± 3.7% 65.9 ± 3.7% 65.9 ± 3.9% 65.9 ± 3.7% 65.9 ± 3.9%
Random forest 65.3 ± 2.4% 65.1 ± 2.4% 65.7 ± 2.4% 65.1 ± 2.4% 65.0 ± 2.5%
GCN 73.2 ± 2.7% 78.7±3.0% 75.8±4.0% 71.7 ± 6.5% 73.4 ± 3.2%
Siamises GCN 59.4 ± 1.7% 58.6 ± 1.9% 60.7 ± 1.2% 62.3 ± 12.0% 61.3 ± 6.9%
ID-GCN(our model) 76.3±3.7% 77.5 ± 4.9% 75 ± 5.9% 81.0±5.5% 77.6±2.8%