|
| 1D-CNN | RNN | LSTM | Transformer | Siam-CNN | Siam-CNN-LSTM |
| Accuracy | 0.9172 | 0.8355 | 0.9359 | 0.9517 | 0.9632 | 0.9655 |
| Precision | 0.8839 | 0.7519 | 0.9016 | 0.9232 | 0.9626 | 0.9510 |
| Recall | 0.8241 | 0.6588 | 0.8784 | 0.9119 | 0.9149 | 0.9308 |
| F1-score | 0.8529 | 0.7023 | 0.8898 | 0.9175 | 0.9381 | 0.9408 |
| AUC | 0.8898 | 0.7840 | 0.9192 | 0.9401 | 0.9496 | 0.9554 |
| 灵敏度/敏感性 | 0.9555 | 0.9092 | 0.9599 | 0.9683 | 0.9844 | 0.9799 |
| 假阳性/误诊率 | 0.1758 | 0.3411 | 0.1215 | 0.0880 | 0.0850 | 0.0691 |
| 假阴性/漏诊率 | 0.0444 | 0.0907 | 0.0400 | 0.0316 | 0.0155 | 0.0200 |