A Quick Negative Selection Algorithm for One-Class Classification in Big Data Era

Table 3

Results in skin segmentation.

Algorithm	DR (%)		FAR (%)		DN		DT (s)		DTT (s)
	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD
OC-SVM	99.09	0.7	51.20	6.67	—	—	—	—	—	—
RNSA	98.42	0.63	0.66	1.48	3000.00	0	8.68	0.13	7501.59	400.49
V-Detector	99.05	0.27	1.31	1.22	469.85	174.66	32.12	23.86	948.55	325.50
BIORV-NSA	99.42	0.34	3.29	2.72	1000.00	0	20.00	0.11	1919.83	59.46
VorNSA	99.20	0.16	1.48	1.49	172.25	11.06	1.91	0.77	671.15	89.36
VorNSA/MR	99.43	0.24	1.56	1.37	176.90	11.96	1.79	0.07	426.70	31.97

The VorNSA/MR is deployed at 2 nodes: one is Intel Pentium [email protected] G (2 Core); the other is Inter Core [email protected] G (2 Core).