Skip to main content

Table 3 Multi-classification analysis with random forest (5-fold cross-validation repeated 100 times independently)

From: Novel biomarker genes which distinguish between smokers and chronic obstructive pulmonary disease patients with machine learning approach

Gene set

Original

Published

Extended

Pred./Truth

NS

SMK

COPD

True rate

NS

SMK

COPD

True rate

NS

SMK

COPD

True rate

NS

25.5

5.8

1.0

0.77

16.1

10.2

4.2

0.48

19.7

8.8

2.4

0.59

SMK

7.2

32.6

15.8

0.76

15.0

25.9

15.0

0.60

13.0

30.4

15.0

0.71

COPD

0.6

4.8

6.7

0.29

2.2

7.0

4.3

0.18

0.6

4.0

6.2

0.26

  1. Classification analysis with random forest was performed using the identified 15 genes (Original) and previously published genes, including genes cited in > 10 (Published) or > 6 publications (Extended)
  2. NS non-smokers, SMK smokers, COPD COPD subjects