Table 2 Results of evaluating model performance using dataset B.
From: Early prediction of postpartum dyslipidemia in gestational diabetes using machine learning models
SVM | Decision tree | Random forest | LightGBM | XGBoost | |
|---|---|---|---|---|---|
Accuracy | 65.21% | 70.16% | 79.77% | 76.66% | 81.05% |
Precision | 65.81% | 68.79% | 81.45% | 76.94% | 81.83% |
Recall | 58.35% | 68.57% | 76.31% | 73.49% | 78.28% |
F1_score | 56.17% | 68.67% | 77.38% | 74.26% | 79.23% |
Specificity | 91.78% | 76.29% | 93.19% | 88.97% | 91.78% |
Sensitivity | 24.91% | 60.85% | 59.43% | 58.01% | 64.77% |
NPV | 64.95% | 74.71% | 77.69% | 76.26% | 79.80% |
PPV | 66.67% | 62.87% | 85.20% | 77.62% | 83.87% |
AUC-ROC | 0.5835 | 0.7153 | 0.8656 | 0.8249 | 0.8792 |
AUC-PR | 0.4645 | 0.6152 | 0.8277 | 0.7762 | 0.8453 |
