Table 2 Summary of performance of machine learning and GNN models
Model | Accuracy | Specificity | Sensitivity | Precision | F1-score |
|---|---|---|---|---|---|
Logistic regression | 62.2 ± 4.2 | 73.6 ± 10.8 | 53.2 ± 10.1 | 67.1 ± 12.5 | 58.0 ± 5.4 |
Support vector machine | 75.5 ± 4.9 | 86.2 ± 7.4 | 65.9 ± 8.9 | 83.1 ± 9.6 | 72.8 ± 6.1 |
K nearest neighbors | 67.6 ± 5.6 | 63.3 ± 8.7 | 72.5 ± 9.8 | 66.3 ± 10.4 | 68.6 ± 7.7 |
AdaBoost | 73.1 ± 7.3 | 66.3 ± 8.5 | 80.2 ± 11.4 | 70.2 ± 10.0 | 74.3 ± 8.8 |
Decision tree | 78.2 ± 6.7 | 72.3 ± 8.3 | 83.3 ± 9.2 | 75.1 ± 9.3 | 78.8 ± 8.6 |
GCN | 77.0 ± 2.3 | 71.4 ± 6.2 | 82.8 ± 6.2 | 74.3 ± 3.6 | 78.1 ± 2.3 |
DGCNN | 80.2 ± 3.3 | 76.4 ± 8.2 | 84.2 ± 5.9 | 77.0 ± 7.7 | 80.2 ± 4.6 |