Skip to main content

Table 4 Median classifiers performance results (and standard deviation in parenthesis) obtained for test sets for the 100 runs tested using five classification methods applied to the fifty DEGs with lowest p-value

From: Identification of biomarkers predictive of metastasis development in early-stage colorectal cancer using network-based regularization

 

Acc

Miscl/FN

Sensitivity

Specificity

AUC

DT

D1

0.61(0.098)

7(1.764)/3(1.465)

0.67(0.163)

0.56(0.167)

0.61(0.092)

D2

0.65(0.103)

6(1.746)/3(1.490)

0.63(0.186)

0.78(0.147)

0.65(0.102)

D3

0.59(0.101)

7(1.717)/3(1.371)

0.63(0.171)

0.67(0.176)

0.59(0.095)

\(\bar{x}\)

0.62

-

0.62

0.67

0.62

svmL

D1

0.67(0.102)

6(1.844)/3(1.581)

0.67(0.176)

0.78(0.209)

0.67(0.102)

D2

0.71(0.092)

5(1.566)/3(1.589)

0.63(0.199)

0.78(0.152)

0.71(0.087)

D3

0.71(0.083)

5(1.415)/4(1.816)

0.50(0.227)

0.89(0.137)

0.69(0.088)

\(\bar{x}\)

0.70

0.60

0.82

0.69

svmR

D1

0.67(0.101)

6(1.817)/3(1.662)

0.67(0.185)

0.56(0.199)

0.67(0.094)

D2

0.59(0.112)

7(1.909)/2(2.567)

0.75(0.321)

0.56(0.222)

0.61(0.114)

D3

0.53(0.090)

8(1.537)/6(1.798)

0.25(0.225)

0.89(0.221)

0.51(0.084)

\(\bar{x}\)

0.60

0.56

0.67

0.60

LR

D1

0.67(0.092)

6(1.663)/3(1.282)

0.67(0.142)

0.67(0.163)

0.67(0.092)

D2

0.65(0.085)

6(1.441)/3(1.299)

0.63(0.162)

0.78(0.132)

0.64(0.082)

D3

0.65(0.105)

6(1.785)/3(1.428)

0.63(0.178)

0.72(0.188)

0.65(0.101)

\(\bar{x}\)

0.66

0.64

0.72

0.65

RF

D1

0.72(0.089)

5(1.602)/3(1.132)

0.72(0.126)

0.78(0.140)

0.72(0.089)

D2

0.71(0.090)

5(1.524)/2(1.329)

0.75(0.166)

0.78(0.140)

0.71(0.091)

D3

0.71(0.102)

5(1.731)/4(1.450)

0.50(0.181)

0.89(0.158)

0.69(0.103)

\(\bar{x}\)

0.71

0.66

0.82

0.71

  1. DT—decision trees; svmL—linear support vector machine; svmR—radial support vector machine; LR—logistic regression; RF—random forest; D1—DATASET1; D2—DATASET2; D3—DATASET3; \(\bar{x}\)—datasets mean; Acc—accuracy; Miscl—misclassifications; FN—false negatives; Sensitivity—fraction of actual positive cases (P); Specificity—fraction of actual negative cases (PM); AUC—area under the ROC curve