Skip to main content

Table 5 Prospective evaluation results of the selected machine learning models

From: Prediction of neonatal deaths in NICUs: development and validation of machine learning models

Model Accuracy Precision Specificity Sensitivity F-score AUC
RF 0.63 0.90 0.72 0.61 0.73 0.81
ANN 0.86 0.96 0.83 0.86 0.91 0.92
C5.0 0.84 0.97 0.89 0.82 0.89 0.91
SVM 0.82 0.94 0.78 0.82 0.88 0.89
Bayesian network 0.67 0.85 0.55 0.71 0.77 0.64
CHAID tree 0.83 0.97 0.89 0.81 0.88 0.91
Ensemble 0.84 0.95 0.83 0.84 0.89 0.91
  1. The best results for each indicator are bold