Skip to main content

Table 7 Median values (and standard deviation in parenthesis) obtained in test set for the five classification methods using fifty most frequently genes pre-selected by LR with EN penalization

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

0.72(0.094)

5(1.698)/2(1.251)

0.78(0.139)

0.67(0.160)

0.72(0.094)

D2

0.65(0.093)

6(1.577)/3(1.282)

0.63(0.160)

0.67(0.168)

0.65(0.087)

D3

0.65(0.119)

6(2.021)/3(1.403)

0.63(0.175)

0.67(0.172)

0.65(0.098)

\(\bar{x}\)

0.67

0.68

0.67

0.67

svmL

D1

0.72(0.081)

5(1.467)/4(1.439)

0.56(0.160)

0.89(0.118)

0.72(0.081)

D2

0.71(0.080)

5(1.355)/3(1.323)

0.63(0.165)

0.89(0.112)

0.70(0.082)

D3

0.71(0.093)

5(1.579)/3(1.463)

0.63(0.183)

0.78(0.130)

0.71(0.091)

\(\bar{x}\)

0.71

0.61

0.85

0.71

svmR

D1

0.78(0.101)

4(1.812)/2(1.043)

0.78(0.116)

0.67(0.196)

0.78(0.101)

D2

0.76(0.095)

4(1.612)/1(0.964)

0.88(0.120)

0.67(0.184)

0.77(0.092)

D3

0.76(0.087)

4(1.480)/2(1.369)

0.75(0.171)

0.78(0.154)

0.76(0.088)

\(\bar{x}\)

0.77

0.80

0.71

0.77

LR

D1

0.72(0.098)

5(1.757)/3(1.255)

0.67(0.139)

0.78(0.156)

0.72(0.098)

D2

0.71(0.092)

5(1.570)/3(1.331)

0.63(0.166)

0.78(0.146)

0.70(0.093)

D3

0.76(0.090)

4(1.535)/2(1.303)

0.75(0.163)

0.78(0.133)

0.76(0.091)

\(\bar{x}\)

0.73

0.68

0.78

0.73

RF

D1

0.78(0.096)

4(1.722)/3(1.143)

0.67(0.127)

0.89(0.127)

0.78(0.096)

D2

0.76(0.078)

4(1.325)/2(1.037)

0.75(0.130)

0.89(0.107)

0.76(0.079)

D3

0.76(0.081)

4(1.384)/3(1.351)

0.63(0.169)

0.89(0.092)

0.76(0.085)

\(\bar{x}\)

0.77

0.68

0.89

0.77

  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