Skip to main content

Table 2 Comparisons of different classifiers : test AUC(std) of breast cancer cases

From: Predicting breast cancer using an expression values weighted clinical classifier

 

Case I

Case II

Case III

Case IV

Case V

Classifiers

     

CL +LS-SVM

     

test AUC

0.7795(0.0687)

0.7772(0.0554)

0.6152(0.0565)

0.6622(0.0628)

0.7740(0.0833)

p-value

0.0039

1.48E-04

0.0086

5.21E-06

0.1602

MA+LS-SVM

     

test AUC

0.7001(0.0559)

0.8065(0.0730)

0.6217(0.0349)

0.7357(0.0085)

0.6166(0.0508)

p-value

0.0059

0.0140

0.0254

2.41E-04

0.0020

GEVD+LS-SVM

     

test AUC

0.7801(0.0717)

0.7673(0.0548)

0.6196(0.0829)

0.7730(0.1011)

0.8001(0.0648)

p-value

0.0137

3.41E-05

0.0040

0.1558

0.0840

KGEVD+LS-SVM

     

test AUC

0.7982(0.0927)

0.8210(0.0670)

0.6437(0.0313)

0.7901(0.0917)

0.8031(0.0624)

p-value

0.0195

0.1144

0.0020

0.6162

0.0720

weighted LS-SVM

     

test AUC

0.8177(0.0666)

0.8465(0.0480)

0.6985(0.0443)

0.8119(0.0893)

0.8210(0.0477)

  1. p-value: a paired test, Wilcoxon signed rank test.
  2. CL and MA are the clinical and microarray kernels of RBF kernel functions.