A hybrid cost-sensitive ensemble for heart disease prediction

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\(^{2}\)	\(0.002^{**}\)	\(0.002^{**}\)	\(0.002^{**}\)	\(0.002^{**}\)	\(0.010^{*}\)	\(0.004^{**}\)
Precision
Z-value	− 2.293	− 2.497	− 1,478	− 2.497	− 1.886	− 2.701
p value	\(0.020^{*}\)	\(0.010^{*}\)	0.160	\(0.010^{*}\)	0.065	\(0.004^{**}\)
Recall
Z-value	− 1.886	− 2.497	− 2.395	− 1.376	− 2.497	− 1.886
p value	0.065	\(0.010^{*}\)	\(0.014^{*}\)	0.193	\(0.010^{*}\)	0.065
G-mean
Z-value	− 2.191	− 2.803	− 2.497	− 2.803	− 1.580	− 2.599
p value	\(0.027^{*}\)	\(0.002^{**}\)	\(0.010^{*}\)	\(0.002^{**}\)	0.131	\(0.006^{**}\)
MC
Z-value	− 2.090	− 2.803	− 2.599	− 2.599	− 2.497	− 2.191
p value	\(0.037^{*}\)	\(0.002^{**}\)	\(0.006^{**}\)	\(0.006^{**}\)	\(0.010^{*}\)	\(0.027^{*}\)
Specificity
Z-value	− 2.599	− 2.497	− 1.886	− 2.803	− 1.988	− 2.803
p value	\(0.006^{**}\)	\(0.010^{*}\)	0.065	\(0.002^{**}\)	\(0.049^{*}\)	\(0.002^{**}\)
AUC
Z-value	− 2.599	− 2.803	− 2.803	− 2.701	− 2.701	− 2.803
p value	\(0.006^{**}\)	\(0.002^{**}\)	\(0.002^{**}\)	\(0.004^{**}\)	\(0.004^{**}\)	\(0.002^{**}\)

ISSN: 1472-6947