Skip to main content

Table 6 Comparison of FS methods applied to both transcriptomics and proteomics, for each combination of classifier and dataset. The number of times each FS method resulted in the highest TPR and the lowest TPR are shown.

From: A machine learning heuristic to identify biologically relevant and minimal biomarker panels from omics data

Method

No. of times the method results in the highest TPR

No. of times the method results in the lowest TPR

CFS

3

15

Chisquared

2

4

NaiveBayes - FS

6

6

Random Forest - FS

10

6

RGIFE

13

5

SVM RFE

16

13