BMC Medical Informatics and Decision Making

Table 5 Performance comparisons of different machine learning models for predicting in-hospital mortality estimated using MIMIC-III test set

From: OASIS +: leveraging machine learning to improve the prognostic accuracy of OASIS severity score for predicting in-hospital mortality

Features	Method	ACC (%)	Sn	Sp	MCC	AUC	Threshold
OASIS subscores	RF200	77.5	0.54	0.80	0.25	0.76	0.16
	LR	73.9	0.66	0.75	0.28	0.78	0.12
	XGB200	70.9	0.79	0.70	0.31	0.81	0.10
OASIS variables	RF200	88.0	0.34	0.94	0.31	0.82	0.33
	LR	70.0	0.69	0.70	0.25	0.77	0.10
	XGB200	72.8	0.78	0.72	0.33	0.83	0.10

Back to article page

ISSN: 1472-6947

Contact us

General enquiries: journalsubmissions@springernature.com