Skip to main content

Table 3 Average classification performance results performing Fivefold cross validation based on the TCGA datasets

From: meth-SemiCancer: a cancer subtype classification framework via semi-supervised learning utilizing DNA methylation profiles

Cancer

Unlabeled dataset

Testing dataset

(pre-training)

Testing dataset

(fine-tuning)

 

Accuracy

F1-score

Accuracy

F1-score

Accuracy

F1-score

BRCA

0.823

0.801

0.800

0.805

0.827

0.817

COAD

0.679

0.678

0.749

0.747

0.757

0.756

GBM

0.921

0.922

0.881

0.884

0.908

0.904

PRAD

0.916

0.901

0.827

0.877

0.891

0.884

RCC

0.964

0.960

0.947

0.949

0.956

0.958

THCA

0.904

0.911

0.842

0.854

0.899

0.894