Table 16.

Wilcoxon signed-rank test: Classifiers with feature subset versus Classifiers with 10 features on Hungarian dataset

	RF	LR	SVM	ELM	KNN	Ensemble
E
Z-value${}^1$	− 2.803	− 2.803	− 2.803	− 2.803	− 2.497	− 2.701
p value${}^2$	$0.{002}^{\ast \ast }$	$0.{002}^{\ast \ast }$	$0.{002}^{\ast \ast }$	$0.{002}^{\ast \ast }$	$0.{010}^{\ast }$	$0.{004}^{\ast \ast }$
Precision
Z-value	− 2.293	− 2.497	− 1,478	− 2.497	− 1.886	− 2.701
p value	$0.{020}^{\ast }$	$0.{010}^{\ast }$	0.160	$0.{010}^{\ast }$	0.065	$0.{004}^{\ast \ast }$
Recall
Z-value	− 1.886	− 2.497	− 2.395	− 1.376	− 2.497	− 1.886
p value	0.065	$0.{010}^{\ast }$	$0.{014}^{\ast }$	0.193	$0.{010}^{\ast }$	0.065
G-mean
Z-value	− 2.191	− 2.803	− 2.497	− 2.803	− 1.580	− 2.599
p value	$0.{027}^{\ast }$	$0.{002}^{\ast \ast }$	$0.{010}^{\ast }$	$0.{002}^{\ast \ast }$	0.131	$0.{006}^{\ast \ast }$
MC
Z-value	− 2.090	− 2.803	− 2.599	− 2.599	− 2.497	− 2.191
p value	$0.{037}^{\ast }$	$0.{002}^{\ast \ast }$	$0.{006}^{\ast \ast }$	$0.{006}^{\ast \ast }$	$0.{010}^{\ast }$	$0.{027}^{\ast }$
Specificity
Z-value	− 2.599	− 2.497	− 1.886	− 2.803	− 1.988	− 2.803
p value	$0.{006}^{\ast \ast }$	$0.{010}^{\ast }$	0.065	$0.{002}^{\ast \ast }$	$0.{049}^{\ast }$	$0.{002}^{\ast \ast }$
AUC
Z-value	− 2.599	− 2.803	− 2.803	− 2.701	− 2.701	− 2.803
p value	$0.{006}^{\ast \ast }$	$0.{002}^{\ast \ast }$	$0.{002}^{\ast \ast }$	$0.{004}^{\ast \ast }$	$0.{004}^{\ast \ast }$	$0.{002}^{\ast \ast }$

${}^1$ The value of Wilcoxon statistics after standardization

${}^2$ ${}^{\ast }$ $p<0.05$, ${}^{\ast \ast }$ $p<0.01$