Table 4.
The performance of supervised learning models on predicting metabolic syndrome and chronic kidney disease.
| Model and disease | Accuracy | Area under the curve (AUC) | F1 score | |
| Classification and regression tree (CART) | ||||
|
|
Metabolic syndrome (Taiwan) | 0.874 | 0.887 | 0.448 |
|
|
Chronic kidney disease (Taiwan) | 0.945 | 0.928 | 0.965 |
| Random forest | ||||
|
|
Metabolic syndrome (Taiwan) | 0.909 | 0.904 | 0.610 |
|
|
Chronic kidney disease (Taiwan) | 0.947 | 0.982 | 0.989 |
|
|
Chronic kidney disease (United States) | 0.951 | 0.929 | 0.679 |
|
|
Chronic kidney disease (Italy) | 0.881 | 0.977 | 0.920 |
|
|
Chronic kidney disease (Japan) | 0.743 | 0.923 | 0.838 |