TABLE 6.

10-fold cross-validation binary mean classification performance for AD against HC using LSVM classifier using different feature selection methods.

Feature selection method	Performance metrics	Mean (%)	Standard deviation	Max (%)	Min (%)
LASSO	Accuracy	90.63	0.515	91.51	88.52
	Sensitivity	87.044	0.585	88.03	85.44
	Specificity	94.315	0.671	95.35	90.95
	F-measure	0.958
FSASL	Accuracy	82.895	1.4020	85.60	80.19
	Sensitivity	78.206	1.5118	81.99	75.21
	Specificity	87.712	1.7666	90.02	82.85
	F-measure	0.8360
LLCFS	Accuracy	81.19	1.438	83.37	77.68
	Sensitivity	85.15	2.087	88.068	80.49
	Specificity	76.39	1.25	78.983	73.65
	F-measure	0.8095
CFS	Accuracy	88.37	1.78	91.18	83.25
	Sensitivity	87.95	1.72	91.98	84.52
	Specificity	88.79	2.17	90.71	81.28
	F-measure	0.903
SVMRFE	Accuracy	65.99	1.48	68.51	62.05
	Sensitivity	65.03	1.41	67.73	61.46
	Specificity	67.40	2.327	70.53	61.27
	F-measure	0.671