Table 7.
Ablation study results on the binary classification task.
| Name | Accuracy | Recall | Precision | F1-Score | AUC |
|---|---|---|---|---|---|
| Binary classification | |||||
| Branch I - small | 0.7810 | 0.8034 | 0.7856 | 0.7944 | 0.9006 |
| Branch II - large | 0.7843 | 0.7986 | 0.7794 | 0.7889 | 0.8991 |
| Dual branch | 0.8310 | 0.8508 | 0.8290 | 0.8397 | 0.9288 |
| Dual branch enhance | 0.7672 | 0.7627 | 0.7430 | 0.7527 | 0.8752 |
| Dual branch attention | 0.8319 | 0.8566 | 0.8386 | 0.8475 | 0.9368 |
| Our proposed | 0.8431 | 0.8667 | 0.8486 | 0.8575 | 0.9357 |
| Multi classification | |||||
| Branch I - small | 0.6764 | 0.7130 | 0.7477 | 0.7299 | 0.8896 |
| Branch II - large | 0.6960 | 0.7366 | 0.7749 | 0.7553 | 0.9148 |
| Dual branch | 0.7385 | 0.7759 | 0.8026 | 0.7891 | 0.9314 |
| Dual branch enhance | 0.7156 | 0.7635 | 0.7925 | 0.7777 | 0.9264 |
| Dual branch attention | 0.7679 | 0.7586 | 0.7895 | 0.7737 | 0.9248 |
| Our proposed | 0.7924 | 0.8422 | 0.8596 | 0.8508 | 0.9635 |