Machine Learning Models for Analysis of Vital Signs Dynamics: A Case for Sepsis Onset Prediction

Table 7

Model performance based on this study’s data with previously presented features.

	LR	SVM-linear	SVM-RBF	SVM- polynomial	ANN
Sensitivity	0.4545	0.7922	0.8442	0.8182	0.6364
Specificity	0.7162	0.3378	0.3919	0.3378	0.5000
PPV	0.6250	0.5545	0.5909	0.5625	0.5698
NPV	0.5579	0.6098	0.7073	0.6410	0.5692
Accuracy	0.5828	0.5695	0.6225	0.5828	0.5695
AUC	0.5914	0.5822	0.6491	0.6018	0.5695

AUC: area under the curve; LR: logistic regression; PPV: positive predictive value; NPV: negative predictive value.