Skip to main content

Table 3 Comparison of maximum recognition accuracies on gene-expression databases using up to 100 top ranked features obtained by different feature-selection methods and a nearest neighbor classifier

From: Ranked selection of nearest discriminating features

  

MRMR [53]

Information gain [54]

Presented

Database

Total number of features

Accuracy (%)

Selected number of features

Accuracy (%)

Selected number of features

Accuracy (%)

Selected number of features

CLL-SUB-111

11340

64.5±6.7

32

64.2±8.0

34

74.9±2.6

1

SMK-CAN-187

19993

65.1±4.3

41

65.1±3.8

29

71.2±1.7

1

GLI-85

22283

83.4±4.5

67

84.2±5.0

87

88.3±2.9

3

GLA-BRA-180

4915

64.8±3.4

45

65.6±4.5

27

68.4±5.1

37

TOX-171

5748

66.2±5.1

100

65.5±5.0

92

69.6±4.4

89