Table 2.
Model performance metrics of 6 models on 10-CV training and validation datasets.
| Model | Sets | AUC | Accuracy | Sensitivity | Specificity | PPV | NPV | F1-score | AP |
|---|---|---|---|---|---|---|---|---|---|
| ExtraRFC | Training Validation | 1.000 0.848 |
0.984 0.795 |
0.951 0.578 |
0.999 0.898 |
0.998 0.730 |
0.977 0.817 |
0.974 0.650 |
0.999 0.743 |
| BernoulliNB | Training Validation | 0.824 0.824 |
0.770 0.771 |
0.649 0.651 |
0.828 0.829 |
0.643 0.645 |
0.832 0.832 |
0.646 0.648 |
0.706 0.707 |
| LogisticReg | Training Validation | 0.837 0.836 |
0.786 0.786 |
0.545 0.546 |
0.901 0.901 |
0.724 0.725 |
0.806 0.806 |
0.622 0.623 |
0.727 0.725 |
| XGBoost | Training Validation | 0.999 0.836 |
0.983 0.783 |
0.956 0.583 |
0.997 0.879 |
0.993 0.698 |
0.979 0.815 |
0.974 0.635 |
0.998 0.722 |
| MLP | Training Validation | 0.871 0.834 |
0.803 0.780 |
0.586 0.551 |
0.906 0.889 |
0.749 0.705 |
0.821 0.806 |
0.657 0.618 |
0.775 0.730 |
| Transformer | Training Validation | 0.842 0.839 |
0.791 0.783 |
0.587 0.573 |
0.888 0.884 |
0.718 0.704 |
0.819 0.812 |
0.644 0.630 |
0.734 0.730 |