TABLE 5.
Dietary biomarkers classifiers on the test set. Precision, recall, F2-score and AUC achieved each algorithm: Decision Tree (DT), Logistic Regression (LR), Naïve Bayes (NB), Neural Network (NN), Random Forest (RF) and Support Vector Machine (SVM), as well as the Bagging and Stacking approaches, using the combinations that achieved the highest F2-score.
| Precision | Recall | F1 | F2 | AUC | |
|---|---|---|---|---|---|
| DT | 0.403 | 0.532 | 0.459 | 0.500 | 0.744 |
| LR | 0.530 | 0.745 | 0.619 | 0.689 | 0.854 |
| NB | 0.515 | 0.745 | 0.609 | 0.684 | 0.853 |
| NN | 0.700 | 0.447 | 0.545 | 0.482 | 0.718 |
| RF | 0.450 | 0.766 | 0.567 | 0.672 | 0.857 |
| SVM | 0.451 | 0.787 | 0.574 | 0.685 | 0.867 |
| Bagging | 0.542 | 0.681 | 0.604 | 0.648 | 0.825 |
| Stacking | 0.388 | 0.851 | 0.533 | 0.687 | 0.889 |
The highest value of each metric is bolded.