Skip to main content

Table 4 Summary of classifiers for the NKI data set with distant metastasis as endpoint

From: Elastic SCAD as a novel penalization method for SVM classification tasks in high-dimensional data

FS method

# features

test error(%)

sensitivity(%)

specificity(%)

Youden index

AUC

L2 SVM

4919 (all)

24

79

68

0.47

0.735

RFE SVM

256

25

83

59

0.42

0.71

MammaPrint(R)

70

37

74

40

0.14

0.57

L1 SVM

1573

17

84

81

0.65

0.825

SCAD SVM

476

25

84

56

0.39

0.695

Elastic Net SVM

109

25

83

59

0.42

0.71

Elastic SCAD SVM

459

24

84

57

0.41

0.705

  1. Misclassification error, sensitivity, specificity, Youden index and AUC value for four feature selection methods, RFE SVM and standard SVM based on ten-fold stratified cross validation.