Table 3 The performance of machine learning models for stillbirth prediction in the test set for nulliparous women. AUC area under the curve, AUPR area under precision-recall curve.
From: Prediction of stillbirth using machine learning methods
Stillbirth type | Period | AUC | AUPR | Specificity | F1 score | Precision | Recall |
|---|---|---|---|---|---|---|---|
All stillbirth | Baseline | 0.707 (0.694–0.719) | 0.217 (0.206–0.229) | 0.828 (0.817–0.838) | 0.044 (0.039–0.050) | 0.024 (0.019–0.028) | 0.404 (0.390–0.417) |
E1 | 0.715 (0.702–0.727) | 0.017 (0.013–0.021) | 0.871 (0.862–0.880) | 0.053 (0.046–0.059) | 0.028 (0.024–0.033) | 0.365 (0.352–0.379) | |
Early stillbirth | Baseline | 0.711 (0.699–0.724) | 0.017 (0.014–0.021) | 0.996 (0.994–0.997) | 0.113 (0.105–0.122) | 0.120 (0.111–0.129) | 0.107 (0.099–0.116) |
E1 | 0.710 (0.698–0.723) | 0.098 (0.090–0.107) | 0.690 (0.678–0.703) | 0.021 (0.017–0.025) | 0.011 (0.008–0.014) | 0.607 (0.594–0.621) | |
Late stillbirth | Baseline | 0.721 (0.709–0.733) | 0.030 (0.025–0.035) | 0.922 (0.915–0.930) | 0.024 (0.020–0.028) | 0.013 (0.009–0.016) | 0.217 (0.206–0.229) |
E1 | 0.670 (0.657–0.683) | 0.007 (0.005–0.010) | 0.943 (0.937–0.949) | 0.019 (0.015–0.023) | 0.010 (0.008–0.013) | 0.130 (0.121–0.140) | |
T0 | 0.742 (0.730–0.754) | 0.013 (0.010–0.016) | 0.962 (0.956–0.967) | 0.045 (0.039–0.051) | 0.025 (0.021–0.029) | 0.217 (0.206–0.229) |
