Skip to main content

Table 5 Percentage accuracy of 10-fold cross validation of feature selection methods applied to the classification methods.

From: A comparative study of different machine learning methods on microarray gene expression data

1. Lymphoma (De vos et.al, 2002)

# Genes

SVM

RBF

MLP

Bayesian

J48

ID3

R. Forest

Bagging

SVM-RFE

50

100.00

96.00

80.00

92.00

88.00

92.00

64.00

64.00

CFS

50

96.00

88.00

76.00

88.00

68.00

84.00

64.00

64.00

ChiSquared

50

96.00

84.00

72.00

92.00

72.00

80.00

80.00

80.00

All features

7129

96.00

84.00

68.00

88.00

64.00

76.00

48.00

52.00

2. Breast (Perou et. al, 2000)

# Genes

SVM

RBF

MLP

Bayesian

J48

ID3

R. Forest

Bagging

SVM-RFE

50

100.00

98.81

97.62

95.24

94.05

97.62

96.43

98.81

CFS

50

98.81

98.81

96.43

94.05

95.24

97.62

97.62

98.81

ChiSquared

50

97.62

97.62

97.62

95.24

95.24

98.81

97.62

98.81

All features

1753

97.62

97.62

96.43

92.86

92.86

96.43

94.05

96.43

3. Colon (Alon et. al, 1999)

# Genes

SVM

RBF

MLP

Bayesian

J48

ID3

R. Forest

Bagging

SVM-RFE

50

100.00

97.78

97.78

97.78

95.56

91.11

95.56

95.56

CFS

50

95.56

93.33

95.56

95.56

95.56

95.56

95.56

97.78

ChiSquared

50

97.78

91.11

95.56

97.78

97.78

95.56

93.33

97.78

All features

7464

95.56

91.11

91.11

93.33

91.11

80.00

88.89

93.33

4. Lung (Garber et. al, 2001)

# Genes

SVM

RBF

MLP

Bayesian

J48

ID3

R. Forest

Bagging

SVM-RFE

50

100.00

98.61

98.61

97.22

97.22

98.61

98.61

98.61

CFS

50

97.22

97.22

98.61

98.61

97.22

97.22

97.22

97.22

ChiSquared

50

98.61

98.61

98.61

97.22

95.83

97.22

97.22

97.22

All features

917

97.22

97.22

97.22

95.83

94.44

95.83

97.22

97.22

5. Adenocarc. (Beer et.al, 2002)

# Genes

SVM

RBF

MLP

Bayesian

J48

ID3

R. Forest

Bagging

SVM-RFE

50

100.00

96.51

94.19

89.53

89.53

89.53

89.53

89.53

CFS

50

98.84

97.67

93.02

89.53

89.53

89.53

89.53

89.53

ChiSquared

50

98.84

95.35

91.86

88.37

90.70

88.37

94.19

88.37

All features

5377

96.51

94.19

75.58

75.58

74.42

79.07

66.28

79.07

6. Lymphoma (Alizadeh et al, 2000)

# Genes

SVM

RBF

MLP

Bayesian

J48

ID3

R. Forest

Bagging

SVM-RFE

50

100.00

100.00

94.79

90.63

90.63

90.63

90.63

90.63

CFS

50

97.92

94.79

93.75

90.63

90.63

90.63

90.63

90.63

ChiSquared

50

97.92

95.83

92.71

89.58

91.67

89.58

94.79

89.58

All features

4027

96.88

88.54

77.08

85.42

75.00

76.04

62.50

84.38

7. Melanoma (Bittner et. al, 2000)

# Genes

SVM

RBF

MLP

Bayesian

J48

ID3

R. Forest

Bagging

SVM-RFE

50

100.00

97.37

92.11

92.11

89.47

89.47

76.32

89.47

CFS

50

97.37

86.84

89.47

89.47

92.11

92.11

76.32

86.84

ChiSquared

50

97.37

89.47

86.84

86.84

89.47

89.47

86.84

86.84

All features

8067

94.74

81.58

84.21

76.32

81.58

81.58

52.63

81.58

8. Ovarian (Welsh et. al, 2001)

# Genes

SVM

RBF

MLP

Bayesian

J48

ID3

R. Forest

Bagging

SVM-RFE

50

100.00

100.00

94.87

94.87

94.87

94.87

92.31

89.74

CFS

50

97.44

87.18

92.31

94.87

94.87

94.87

89.74

92.31

ChiSquared

50

97.44

89.74

92.31

94.87

94.87

94.87

87.18

94.87

All features

7129

94.87

84.62

89.74

87.18

87.18

89.74

74.36

89.74