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

Table 8

Excellent performance metrics results and best SVM kernel on selected feature subset.

Predictive model	Best feature subset	Accuracy (%)	Specificity (%)	Sensitivity/recall (%)	F1-score	MCC	Error	Execution time
SVM (kernel-linear)	18	99	99	98	99	99	1	0.030
SVM (kernel = RBF)	18	98	99	98	98	97	2	0.004
SVM (kernel = polynomial)	18	97	97	97	97	97	3	0.002
SVM (kernel = sigmoid)	13	84	54	60	45	77	16	0.005