Skip to main content

Table 6 Comparison of performance for the CNS and Lung Cancer data sets using six classifiers with the same number of genes chosen by three gene selection methods

From: Discovery of dominant and dormant genes from expression data using a novel generalization of SNR for multi-class problems

Data Sets

Gene Selection Methods

Classifiers

m-fold of genes

   

1-fold

2-fold

3-fold

4-fold

5-fold

CNS

GDI.Dominant

OVO.SVM-L

35.5 ± 1.0

32.0 ± 1.1

27.4 ± 1.0

26.6 ± 1.1

24.3 ± 1.0

  

OVO.SVM-R

36.5 ± 1.1

30.8 ± 1.1

27.4 ± 1.1

25.8 ± 1.1

24.0 ± 1.0

  

OVA.SVM-L

35.9 ± 1.4

32.8 ± 1.2

28.8 ± 1.0

27.8 ± 1.0

26.3 ± 1.0

  

OVA.SVM-R

35.2 ± 1.2

29.9 ± 1.2

27.7 ± 1.0

26.3 ± 1.0

25.3 ± 1.0

  

NMC

33.5 ± 1.1

27.0 ± 1.0

23.7 ± 1.0

22.9 ± 1.0

21.4 ± 0.9

  

NNC

33.0 ± 1.1

28.1 ± 1.1

25.6 ± 1.0

26.1 ± 0.9

25.1 ± 0.9

 

OVA.SNR [12]

OVO.SVM-L

37.1 ± 1.1

30.5 ± 1.0

27.0 ± 1.0

23.5 ± 1.0

21.6 ± 1.0

  

OVO.SVM-R

35.9 ± 1.1

29.6 ± 1.0

27.1 ± 1.0

23.2 ± 1.0

21.6 ± 1.0

  

OVA.SVM-L

36.8 ± 1.1

30.9 ± 1.1

26.8 ± 0.9

22.8 ± 0.9

20.8 ± 0.9

  

OVA.SVM-R

35.2 ± 1.1

29.1 ± 0.9

26.1 ± 0.9

23.7 ± 0.9

21.6 ± 1.0

  

NMC

32.8 ± 1.1

26.3 ± 1.0

23.5 ± 1.0

20.6 ± 0.8

18.5 ± 0.9

  

NNC

34.9 ± 1.0

28.3 ± 1.0

26.0 ± 1.0

24.0 ± 0.9

21.5 ± 1.0

 

ANOVA+Correlation [11]

OVO.SVM-L

38.5 ± 1.1

32.2 ± 1.0

27.0 ± 0.9

24.0 ± 0.8

21.6 ± 0.8

  

OVO.SVM-R

37.1 ± 1.1

30.8 ± 1.0

27.4 ± 1.0

25.0 ± 0.9

21.4 ± 0.8

  

OVA.SVM-L

38.5 ± 1.1

33.4 ± 1.0

27.4 ± 1.0

25.5 ± 0.8

23.2 ± 0.8

  

OVA.SVM-R

35.8 ± 1.1

31.0 ± 1.1

26.0 ± 1.0

25.1 ± 0.8

24.1 ± 0.9

  

NMC

33.7 ± 1.3

24.9 ± 0.9

20.8 ± 0.8

19.7 ± 0.8

19.2 ± 0.8

  

NNC

36.0 ± 1.2

29.7 ± 1.0

24.4 ± 0.9

23.3 ± 0.8

21.4 ± 0.7

Lung Cancer

GDI.Dominant

OVO.SVM-L

9.5 ± 0.3

8.1 ± 0.3

7.8 ± 0.3

7.8 ± 0.3

7.4 ± 0.3

  

OVO.SVM-R

10.0 ± 0.4

8.3 ± 0.3

8.1 ± 0.3

7.1 ± 0.3

6.9 ± 0.3

  

OVA.SVM-L

9.4 ± 0.3

7.7 ± 0.3

7.8 ± 0.3

7.5 ± 0.3

7.8 ± 0.3

  

OVA.SVM-R

10.1 ± 0.3

8.3 ± 0.3

8.2 ± 0.3

7.5 ± 0.3

7.4 ± 0.3

  

NMC

9.8 ± 0.4

7.2 ± 0.3

6.3 ± 0.2

5.8 ± 0.2

5.8 ± 0.3

  

NNC

11.9 ± 0.4

9.0 ± 0.3

8.3 ± 0.3

7.7 ± 0.2

7.3 ± 0.3

 

OVA.SNR [12]

OVO.SVM-L

9.8 ± 0.3

8.0 ± 0.3

8.1 ± 0.3

8.0 ± 0.3

7.4 ± 0.3

  

OVO.SVM-R

10.2 ± 0.3

8.8 ± 0.3

7.6 ± 0.3

7.4 ± 0.3

7.2 ± 0.2

  

OVA.SVM-L

9.6 ± 0.3

8.3 ± 0.3

7.9 ± 0.3

7.9 ± 0.3

7.9 ± 0.3

  

OVA.SVM-R

10.0 ± 0.3

8.7 ± 0.3

8.1 ± 0.3

7.7 ± 0.3

7.2 ± 0.3

  

NMC

9.5 ± 0.3

7.6 ± 0.3

6.7 ± 0.2

6.5 ± 0.2

6.1 ± 0.2

  

NNC

11.9 ± 0.3

9.3 ± 0.3

7.8 ± 0.2

7.3 ± 0.2

7.4 ± 0.3

 

ANOVA+Correlation [11]

OVO.SVM-L

6.9 ± 0.3

7.5 ± 0.3

7.4 ± 0.3

7.5 ± 0.3

7.7 ± 0.3

  

OVO.SVM-R

7.6 ± 0.3

7.1 ± 0.3

6.4 ± 0.3

6.6 ± 0.2

6.7 ± 0.3

  

OVA.SVM-L

7.1 ± 0.3

6.9 ± 0.3

7.1 ± 0.3

8.1 ± 0.3

7.8 ± 0.3

  

OVA.SVM-R

7.9 ± 0.3

7.4 ± 0.3

7.1 ± 0.3

7.4 ± 0.3

6.7 ± 0.3

  

NMC

7.8 ± 0.3

6.3 ± 0.3

5.7 ± 0.2

5.1 ± 0.2

5.3 ± 0.2

  

NNC

9.8 ± 0.3

8.0 ± 0.3

7.5 ± 0.3

7.3 ± 0.3

6.7 ± 0.3

  1. Here m-fold corresponds to the case when m top most dominant genes are used for each class. For example, the column labeled 3-fold represents the results using 15 genes (3 dominant genes from each of the 5 classes) for both the CNS and Lung Cancer data sets.