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

Table 4

Classification results of SVM (kernel = linear).

Model	Performance evaluation metric
	Parameter (C, )	No of features	Acc (%)	Sp (%)	Precision (%)	Sen/recall (%)	F1-score	MCC	Classification error (%)	Execution time (s)
SVM (linear)	(1, 0.0001)	1	76	88	88	56	70	72	24	0.003
		2	85	90	92	74	82	83	15	0.083
		3	95	94	94	97	96	95	5	0.002
		4	96	94	94	95	95	95	4	0.005
		5	96	96	96	94	95	95	4	0.003
		6	96	96	93	94	95	96	4	0.002
		7	96	96	96	96	96	94	4	0.002
		8	95	94	94	94	94	95	5	0.002
		9	96	96	96	95	94	96	4	0.002
		10	97	96	95	94	95	96	3	0.004
		11	97	96	96	94	95	96	3	0.003
		12	97	97	97	98	97	96	3	0.010
		13	95	96	93	94	95	96	5	0.010
		14	95	97	97	94	96	96	5	0.010
		15	98	99	99	97	98	98	2	0.024
		16	98	99	99	97	98	98	2	0.019
		17	98	98	98	98	98	98	2	0.080
		18	99	99	99	98	99	99	1	0.030
		19	97	96	96	99	98	97	3	0.022
		20	97	98	98	98	98	97	3	0.027
		21	97	96	96	99	98	97	3	0.038
		22	96	96	96	99	98	97	4	0.107
		23	98	98	98	99	98	98	2	1.832
		24	98	98	98	99	98	98	2	0.909
		25	97	96	96	97	97	97	3	0.733
		26	96	96	96	97	97	97	4	3.729
		27	96	95	95	97	96	96	4	3.987
		28	96	95	95	96	95	96	4	3.594
		29	96	97	97	97	97	97	4	4.332
		30	95	96	96	95	98	94	5	4.547