BMC Medical Informatics and Decision Making

Table 5 Performance in stratified subgroups

From: Risk prediction of heart failure in patients with ischemic heart disease using network analytics and stacking ensemble learning

Subgroup	Metrics [Mean ± SD]
Subgroup	Precision	Recall	Accuracy	F₁ score	AUC
Sex
Male	0.716 ± 0.020	0.881 ± 0.017	0.848 ± 0.010	0.790 ± 0.015	0.926 ± 0.007
Female	0.722 ± 0.018	0.875 ± 0.019	0.857 ± 0.009	0.791 ± 0.014	0.932 ± 0.006
P-value^a	0.0327	0.0277	< 0.0001	0.5533	< 0.0001
Age groups
18–44	0.798 ± 0.185	0.901 ± 0.143	0.960 ± 0.030	0.832 ± 0.148	0.975 ± 0.044
45–59	0.685 ± 0.053	0.801 ± 0.060	0.919 ± 0.014	0.736 ± 0.042	0.953 ± 0.013
60–69	0.658 ± 0.034	0.809 ± 0.029	0.860 ± 0.013	0.725 ± 0.026	0.924 ± 0.010
70–79	0.725 ± 0.022	0.887 ± 0.018	0.838 ± 0.012	0.798 ± 0.015	0.921 ± 0.008
80+	0.757 ± 0.021	0.928 ± 0.017	0.816 ± 0.014	0.834 ± 0.015	0.894 ± 0.013
P-value^b	< 0.0001	< 0.0001	< 0.0001	< 0.0001	< 0.0001

^a: t-test; ^b: one-way analysis of variance.

Back to article page

ISSN: 1472-6947

Contact us

General enquiries: journalsubmissions@springernature.com