Table 5 . Performance metrics of the SVM classifiers constructed with the four feature sets formed from different combinations of the selected number of principal components.

	Accuracy (std)	AUC (std)	Sensitivity (std)	Specificity (std)	Precision (std)	F1 score (std)	Time (s) (std)
Feature set (1) (250 features)	93.4% (0.002)	0.98 (0.001)	0.945 (0.006)	0.922 (0.003)	0.919 (0.004)	0.933 (0.003)	2.071 (0.269)
Feature set (2) (300 features)	94% (0.002)	0.98 (0.001)	0.949 (0.001)	0.932 (0.001)	0.93 (0.001)	0.94 (0.001)	2.105 (0.001)
Feature set (3) (350 features)	92.6% (0.003)	0.972 (0.005)	0.935 (0.006)	0.916 (0.005)	0.913 (0.005)	0.924 (0.004)	2.002 (0.194)
Feature set (4) (300 features)	93% (0.003)	0.98 (0.001)	0.935 (0.005)	0.926 (0.005)	0.925 (0.006)	0.93 (0.001)	2.088 (0.293)

Note:

Bold values indicate the highest results.