Skip to main content

Table 3 Diagnosis accuracy, sensitivity and specificity based on several classification methods with fivefold cross-validation

From: Identification and validation of the methylation biomarkers of non-small cell lung cancer (NSCLC)

 

Test

Train

 

Sensitivity

Specificity

Accuracy

Sensitivity

Specificity

Accuracy

Logistic regression

0.791

0.993

0.891

0.775

0.969

0.871

SVMa

0.897

0.977

0.937

0.855

0.941

0.897

Random forest

0.934

0.928

0.931

0.890

0.886

0.886

Bayes tree

0.911

0.976

0.944

0.863

0.957

0.909

  1. aSVM represents support vector machines and Kernel Methods. Sensitivity, specificity and classification accuracy were its mean value in fivefold validations with 1,000 replications. In the main body of the manuscript, sensitivity, specificity and accuracy were derived from training result of the classification.