BMC Medical Research Methodology

Table 15 Attribute disclosure performance metric when the attacker wants for uncover 5 unknown attributes. BREAST large-set

From: Generation and evaluation of synthetic patient data

	# of synthetic samples
	5k		10k		20k		∼170k (all)
	k=1	k=10	k=1	k=10	k=1	k=10	k=1	k=10
IM	0.931	0.876	0.946	0.886	0.955	0.887	0.971	0.883
BN	0.947	0.932	0.954	0.942	0.967	0.945	0.982	0.946
MPoM	0.952	0.945	0.964	0.950	0.967	0.949	0.981	0.951
CLGP	0.958	0.950	0.968	0.949	0.974	0.950	0.987	0.953
MC-MedGAN	0.024	0.024	0.024	0.024	0.024	0.024	0.024	0.024
MICE-LR	0.948	0.936	0.960	0.944	0.966	0.948	0.986	0.959
MICE-LR-DESC	0.954	0.946	0.963	0.946	0.967	0.950	0.984	0.955
MICE-DT	0.959	0.952	0.973	0.954	0.983	0.957	0.998	0.972

Back to article page

ISSN: 1471-2288

Contact us

General enquiries: journalsubmissions@springernature.com