Table 4.

Comparative experimental results of the best subsets produced by mABC and other methods for different datasets

Dataset name	Evaluation	EPSO				IBPSO		TS-BPSO		BPSO-CGA		Random forest			mABC
	Criteria	[37]				[43]		[42]		[44]		[55]			[This work]
		Best	Avg.	S.D.		Best		Best		Best		Avg.	S.D.		Best	Avg.	S.D.
9_Tumors	Accuracy	76.67	75.00	1.11		78.33		81.63		-		88.0	0.037		100	98.65	0.01
	# Genes	251	247.10	9.65		1280		2941		-		130	-		30	34.73	5.64
11_Tumors	Accuracy	96.55	95.40	0.61		93.10		97.35		-		-	-		100	99.50	0.00
	# Genes	243	237.70	9.66		2948		3206		-		-	-		42	47.27	7.79
Brain_Tumor1	Accuracy	93.33	92.11	0.82		94.44		95.89		91.4		93.6	0.007		100	100	0.0
	# Genes	8	7.5	2.51		754		2913		456		86	-		12	16.87	2.85
Brain_Tumor2	Accuracy	94	92.4	1.27		94.00		92.65		-		-	-		100	100	0.0
	# Genes	4	6.0	1.83		1197		5086		-		-	-		7	10.52	1.72
DLBCL	Accuracy	100	100	0		100		100		-		94.6	0.021		100	100	0.0
	# Genes	3	4.7	0.82		1042		2671		-		21	-		3	4.05	0.78
Leukemia1	Accuracy	100	100	0		100		100		100		98.1	0.006		100	100	0.0
	# Genes	2	3.20	0.63		1034		2577		300		21	-		4	5.67	0.73
Leukemia2	Accuracy	100	100	0		100		100		-		-	-		100	100	0.0
	# Genes	4	6.8	2.2		1292		5609		-		-	-		4	6.29	0.98
Lung_Cancer	Accuracy	96.06	95.67	0.31		96.55		99.52		-		-	-		100	100	0.0
	# Genes	7	8.3	2.11		1897		6958		-		-	-		14	23.31	5.14
Prostate_Tumor	Accuracy	99.02	97.84	0.62		92.16		95.45		93.7		-	-		100	100	0.0
	# Genes	5	6.6	2.17		1294		5320		795		-	-		5	10.73	3.15
SRBCT	Accuracy	100	99.64	0.58		100		100		100		99.7	0.003		100	100	0.0
	# Genes	7	14.90	13.03		431		1084		880		42	-		5	5.59	0.51

Best results (maximum accuracy and minimum selected gene size) are highlighted using boldface font