Table 2.

Comparison of the classification performances of the supervised machine learning accuracies between the 22-features and the 7-features models.

		SVM		RF		DT		KNN		MLP-ANN
		22 Features	7 Features	22 Features	7 Features	22 Features	7 Features	22 Features	7 Features	22 Features	7 Features
Accuracy Training	Mean10runs (SD)	0.91 (0.01)	0.88 (0.01)	0.89 (0.01)	0.87 (0.01)	0.89 (0.02)	0.85 (0.01)	0.86 (0.03)	0.82 (0.02)	0.87 (0.02)	0.84 (0.03)
	p (d)	0.00 (2.48)		0.00 (1.59)		0.00 (2.13)		0.00 (1.59)		0.03 (1.03)
Accuracy Test	Mean10runs (SD)	0.81 (0.01)	0.86 (0.02)	0.79 (0.01)	0.82 (0.01)	0.76 (0.02)	0.79 (0.01)	0.68 (0.02)	0.74 (0.02)	0.74 (0.02)	0.77 (0.03)
	p (d)	0.00 (3.17)		0.00 (2.68)		0.00 (1.79)		0.00 (2.99)		0.01 (1.32)
Precision	Mean10runs (SD)	0.80 (0.01)	0.85 (0.02)	0.79	0.83	0.76 (0.03)	0.81 (0.02)	0.68 (0.02)	0.75 (0.02)	0.74 (0.03)	0.78 (0.03)
	p (d)	0.00 (3.12)		0.00 (2.94)		0.00 (1.75)		0.00 (3.03)		0.00 (1.19)
Recall	Mean10runs (SD)	0.80 (0.02)	0.85 (0.03)	0.78 (0.03)	0.82 (0.01)	0.75 (0.03)	0.79 (0.02)	0.69 (0.02)	0.74 (0.03)	0.74 (0.02)	0.77 (0.03)
	p (d)	0.00 (2.42)		0.01 (1.45)		0.00 (1.67)		0.00 (2.08)		0.00(1.59)
F1score	Mean10runs (SD)	0.80 (0.01)	0.85 (0.02)	0.78 (0.03)	0.82 (0.01)	0.75 (0.03)	0.80 (0.02)	0.68 (0.02)	0.74 (0.02)	0.74 (0.03)	0.78 (0.02)
	p (d)	0.00 (2.65)		0.00 (2.01)		0.00 (1.88)		0.00 (2.58)		0.02 (1.54)
AUC	Mean10runs (SD)	0.80 (0.01)	0.85 (0.02)	0.78 (0.03)	0.82 (0.01)	0.75 (0.03)	0.79 (0.01)	0.68 (0.02)	0.74 (0.02)	0.73 (0.03)	0.77 (0.02)
	p (d)	0.00 (2.75)		0.00 (1.86)		0.00 (1.85)		0.00 (2.43)		0.00 (1.56)
Generalization Error (%)	Mean10runs (SD)	9.35 (1.31)	2.95 (1.46)	9.93 (1.16)	4.96 (0.92)	12.44 (1.31)	5.47 (0.91)	17.70 (0.69)	7.84 (2.02)	12.80 (2.51)	7.26 (0.92)
	p (d)	0.00 (4.62)		0.00 (4.71)		0.00 (6.17)		0.00 (6.53)		0.00 (2.93)

SVM, support vector machine; RF, random forest; DT, decision tree; KNN, k-nearest neighbor; MLP-ANN, multilayer perceptron artificial neural networks; SD, standard deviation; d, Cohen’s effect size.