Table 7.

Results from five simulations, including average and standard deviation (SD), after deep features from the VGG16 model are extracted and fed to several FS algorithms with embedded local search

FS method	LAHC embedded FS method		ABHC embedded FS method
	Classification accuracy (%)	No. of features selected	Classification accuracy (%)	No. of features selected
GSA	93.30	65	95.08	70
	94.28	54	94.19	63
	94.19	51	94.28	56
	93.30	67	92.41	57
	93.33	65	93.30	64
Mean ± SD	93.68 ± 0.51	54 ± 7	93.85 ± 0.92	62 ± 6
WOA	94.64	48	94.64	55
	93.33	58	93.75	58
	93.75	59	93.33	61
	94.19	68	94.64	69
	94.19	48	94.64	60
Mean ± SD	94.02 ± 0.50	56 ± 8	94.19 ± 0.55	61 ± 5
GWO	94.64	76	94.64	69
	93.33	65	93.30	78
	93.75	81	93.33	63
	94.64	74	93.75	79
	93.33	65	94.64	75
Mean ± SD	93.94 ± 0.67	72 ± 7	93.93 ± 0.59	73 ± 7
GA	94.64	70	94.64	67
	93.30	59	93.75	62
	93.30	57	93.75	55
	93.33	56	93.33	59
	94.64	70	93.30	59
Mean ± SD	93.84 ± 0.73	62 ± 7	93.75 ± 0.48	60 ± 4
PSO	93.30	41	94.64	55
	93.33	42	94.19	52
	94.19	50	93.30	51
	92.86	46	93.33	48
	93.33	52	94.64	57
Mean ± SD	93.40 ± 0.48	46 ± 5	94.02 ± 0.59	53 ± 4
SCA	93.30	59	94.19	70
	93.33	71	93.75	67
	94.19	73	93.75	64
	93.33	65	93.33	63
	93.41	57	94.19	70
Mean ± SD	93.51 ± 0.38	65 ± 7	93.84 ± 0.32	67 ± 3
HS	93.30	63	92.41	60
	94.28	63	91.16	73
	95.08	56	93.33	53
	93.33	55	91.52	62
	92.42	65	91.96	68
Mean ± SD	93.68 ± 1.02	60 ± 5	92.07 ± 0.75	63 ± 8
EO	93.75	53	94.64	55
	94.28	56	94.64	69
	94.64	56	93.30	55
	94.64	50	94.28	63
	93.41	55	93.33	55
Mean ± SD	94.14 ± 0.55	54 ± 3	94.04 ± 0.60	59 ± 6