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

Table 7

Classification results of SVM (kernel = sigmoid).

Model	Classification performances evaluation metrics
	Parameters (C, )	Number of features	Acc (%)	Sp (%)	Precision (%)	Sen/recall (%)	F1-score	MCC	Classification error	Execution time (s)
SVM (sigmoid)	(1, 0.0001)	1	64	100	99	20	34	50	36	0.006
		2	64	100	99	20	34	50	36	0.006
		3	71	100	99	20	34	60	29	0.006
		4	79	100	99	42	56	70	21	0.006
		5	78	100	99	62	77	70	22	0.011
		6	79	100	99	42	59	71	21	0.005
		7	79	100	100	63	77	71	21	0.005
		8	80	100	100	43	60	72	20	0.005
		9	80	100	100	42	59	71	20	0.008
		10	80	100	100	42	59	71	20	0.005
		11	81	100	100	42	59	70	19	0.006
		12	80	100	100	41	58	50	20	0.016
		13	84	54	54	60	45	77	16	0.005
		14	75	99	99	70	82	60	25	0.009
		15	76	100	100	72	83	63	24	0.010
		16	74	100	100	64	77	64	26	0.008
		17	45	70	70	01	2	35	55	0.031
		18	28	45	45	2	4	22	72	0.012
		19	28	45	45	02	4	22	72	0.020
		20	28	45	45	02	4	22	72	0.011
		21	29	45	45	02	4	22	71	0.148
		22	28	45	45	02	4	22	72	0.044
		23	28	45	45	02	4	22	72	0.044
		24	28	45	45	02	4	22	72	0.004
		25	28	45	45	02	4	22	72	0.004
		26	27	45	45	02	4	22	73	0.014
		27	27	45	45	02	4	22	73	0.016
		28	27	45	45	02	4	22	73	0.017
		29	27	45	45	02	4	22	73	0.117
		30	27	45	45	02	4	22	73	0.221