Breast Cancer Detection with Reduced Feature Set

Table 3

The confusion matrices of the classifiers using reduced one dimensionality by ICA (1F denotes one feature and 30F denotes original features).

-NN classifier ()					ANN classifier (neuron number 7)
	Recognized value					Recognized value
Actual value	Malignant		Benign		Actual value	Malignant		Benign
	1F	30F	1F	30F		1F	30F	1F	30F
Malignant	338 (TP)	346	19 (FN)	11	Malignant	346 (TP)	357	11 (FN)	0
Benign	32 (FP)	28	180 (TN)	184	Benign	43 (FP)	14	169 (TN)	198
RBFNN classifier (spread = 28)					SVM classifier ()
	Recognized value					Recognized value
Actual value	Malignant		Benign		Actual value	Malignant		Benign
	1F	30F	1F	30F		1F	30F	1F	30F
Malignant	345 (TP)	334	12 (FN)	23	Malignant	348 (TP)	343	14 (FN)	9
Benign	43 (FP)	138	169 (TN)	74	Benign	43 (FP)	13	169 (TN)	199