Skip to main content

Table 2 Minimum classification error rate estimated for each data set for the first best approaches (percentage) and the associated number of genes/SNPs that were selected.

From: Sparse PLS discriminant analysis: biologically relevant feature selection and graphical displays for multiclass problems

Data set

rank 1

rank 2

rank 3

rank 4

rank 5

rank 6

rank 7

rank 8

rank 9

Leukemia

RFE

SPLSDA-

LDA

LDA

SPLSDA-

LOG

RF

DDA

sPLS

NSC

SGPLS

error rate

20.55

22.36

22.78

23.33

24.17

24.31

24.30

26.25

26.67

# genes

5

200

7129

500

200

50

10

500

500

SRBCT

RF

OFW-

cart

DDA

LDA

sPLS

NSC

SGPLS

RFE

SPLSDA-

LDA

error rate

0.00

0.00

0.00

0.00

0.16

0.63

1.27

1.58

1.90

# genes

30

50

30

100

100

500

50

5

200

Brain

RFE

DDA

LDA

sPLS

RF

SPLSDA-

LDA

NSC

OFW-

cart

SPLSDA-

LOG

error rate

10.56

10.78

11.11

11.22

11.89

14.45

15.11

15.56

17.00

# genes

10

25

30

6144

500

35

20

35

50

GCM

RFE

LDA

RF

SGPLS-

LDA

sPLS

OFW-

svm

SGPLS-

LOG

OFW-

cart

NSC

error rate

0.81

1.14

1.22

1.63

3.41

4.01

4.71

4.88

7.23

# genes

5

500

500

200

200

500

500

7129

10

SNP

NSC

DDA

SPLS

RFE

SPLSDA-

LDA

RF

SPLSDA-

LOG

OFW-

cart

OFW-

svm

error rate

6.50

11.54

11.71

12.36

13.01

17.40

31.22

49.96

51.67

# SNPs

5000

1000

2000

20000

2000

20000

200

20000

20000

  1. The approaches are ranked by their performances.