Table 6 Model Results using the top 26 variables selected from 37 variables producing the highest F1-score
From: Machine learning based model for the early detection of Gestational Diabetes Mellitus
Model | Sn (Recall) | Sp | Acc | F1-score | Precision |
|---|---|---|---|---|---|
RandomForestClassifier | 81.72% | 78.51% | 80.22% | 81.08% | 81.55% |
GradientBoostingClassifier | 82.85% | 77.69% | 80.11% | 81.31% | 80.83% |
AdaBoostClassifier | 85.67% | 78.24% | 81.94% | 83.27% | 82.15% |
DecisionTreeClassifier | 79.16% | 78.51% | 78.78% | 79.61% | 81.35% |
LogisticRegression | 80.55% | 77.44% | 79.22% | 79.18% | 78.53% |
SVC | 80.34% | 75.97% | 78.06% | 79.29% | 79.58% |
GaussianNB | 82.81% | 70.23% | 77.28% | 78.78% | 76.53% |
KNeighborsClassifier | 88.29% | 67.25% | 78.39% | 81.14% | 75.80% |
CatBoostClassifier | 89.22% | 73.62% | 81.56% | 83.45% | 79.34% |
XGBClassifier | 78.62% | 79.04% | 78.61% | 79.18% | 80.93% |
LGBMClassifier | 73.85% | 71.92% | 73.39% | 73.38% | 73.82% |
Stacking | 92.13% | 84.94% | 88.83% | 89.56% | 87.34% |