Breast Cancer Detection in the IOT Health Environment Using Modified Recursive Feature Selection

Table 6

Classification results of SVM (kernel = polynomial).

Model	Classification performances evaluation metrics
	Parameters (C, )	Number of features in subset	Acc (%)	Sp (%)	Precision (%)	Sen/recall (%)	F1-score	MCC	Classification error	Execution time (s)
SVM (polynomial)	(1, 0.0001)	1	64	100	99	20	33	50	36	0.013
		2	65	99	98	23	37	50	35	0.009
		3	63	99	98	23	37	50	37	0.005
		4	63	99	98	23	37	50	37	0.006
		5	63	99	99	23	37	50	37	0.006
		6	64	99	99	23	37	50	36	0.006
		7	64	99	99	23	37	51	36	0.004
		8	65	99	99	24	39	51	35	0.005
		9	64	99	99	24	39	50	36	0.007
		10	64	99	99	24	39	50	36	0.006
		11	65	99	99	24	39	51	35	0.006
		12	90	99	99	75	95	87	10	0.002
		13	89	99	99	40	57	80	11	0.004
		14	88	99	99	74	85	86	12	0.002
		15	92	99	99	81	89	91	8	0.002
		16	93	98	98	81	88	91	7	0.002
		17	93	98	98	87	92	92	7	0.005
		18	97	97	97	97	97	97	3	0.002
		19	96	97	97	97	97	97	4	0.119
		20	96	97	97	95	96	97	4	0.111
		21	96	96	96	96	96	97	4	0.124
		22	96	96	96	96	96	97	4	0.060
		23	95	95	95	96	95	95	5	0.060
		24	94	94	94	96	94	93	6	0.061
		25	93	94	94	95	94	93	7	0.151
		26	92	95	95	95	94	93	8	0.162
		27	92	94	94	94	93	92	8	0.167
		28	92	94	94	94	93	92	8	0.211
		29	92	94	94	91	92	91	8	0.234
		30	92	92	92	91	91	92	8	0.277