Table 4 External Validation with feature selection.
From: Boosting pre-trained model with silica nanoparticles cellular toxicity prediction
Method | Accuracy | Precision | Recall | F1-score | AUC-ROC |
|---|---|---|---|---|---|
GradientBoost | 0.7856 | 0.7184 | 0.7399 | 0.7271 | 0.8466 |
RandomForest | 0.7834 | 0.7556 | 0.5912 | 0.5976 | 0.8174 |
LogisticRegression | 0.6718 | 0.5792 | 0.5868 | 0.5817 | 0.6551 |
XGBoost | 0.8033 | 0.7396 | 0.7621 | 0.7490 | 0.8529 |
SVC | 0.7635 | 0.6801 | 0.5661 | 0.5631 | 0.6797 |
BPNN | 0.7238 | 0.6684 | 0.7091 | 0.6755 | 0.7864 |
CatBoost | 0.7934 | 0.7286 | 0.7540 | 0.7387 | 0.8683 |
FTTransformer | 0.6862 | 0.6106 | 0.6291 | 0.6151 | 0.6728 |
TabTransformer | 0.7576 | 0.7362 | 0.7210 | 0.7268 | 0.7210 |
TabPFN | 0.8564 | 0.8169 | 0.7825 | 0.7971 | 0.8938 |
