Logistic regression model training based on the approximate homomorphic encryption

BMC Medical Genomics

Table 2 Implementation results for other datasets with 5-fold CV

Dataset	Sample	Feature	Method	degg	Iter	Enc	Learn	Storage	Accuracy	AUC
	num	num			num	time	time
Edinburgh	1253	9	Ours	5	7	2s	3.6 min	0.02 GB	91.04%	0.958
			[13]	3	25	12s	114 min	0.69 GB	86.03%	0.956
			[13]	7	20	12s	114 min	0.71 GB	86.19%	0.954
lbw	189	9	Ours	5	7	2s	3.3 min	0.02 GB	69.19%	0.689
			[13]	3	25	11s	99 min	0.67 GB	69.30%	0.665
			[13]	7	20	11s	86 min	0.70 GB	69.29%	0.678
nhanes3	15649	15	Ours	5	7	14s	7.3 min	0.16 GB	79.22%	0.717
			[13]	3	25	21s	235 min	1.15 GB	79.23%	0.732
			[13]	7	20	21s	208 min	1.17 GB	79.23%	0.737
pcs	379	9	Ours	5	7	2s	3.5 min	0.02 GB	68.27%	0.740
			[13]	3	25	11s	103 min	0.68 GB	68.85%	0.742
			[13]	7	20	11s	97 min	0.70 GB	69.12%	0.750
uis	575	8	Ours	5	7	2s	3.5 min	0.02 GB	74.44%	0.603
			[13]	3	25	10s	104 min	0.61 GB	74.43%	0.585
			[13]	7	20	10s	96 min	0.63 GB	75.43%	0.617

ISSN: 1755-8794