Skip to main content

Table 2 Performance of the SVM using different feature subsets selected by Info Gain

From: BLProt: prediction of bioluminescent proteins based on support vector machine and relieff feature selection

Feature subset

Sensitivity

(%)

Specificity

(%)

MCC

Test Accuracy (%)

CV

Accuracy

100 features

69.50

74.21

0.4351

72.21

74.83

200 features

76.60

75.79

0.5193

76.13

78.00

300 features

70.92

77.37

0.4821

74.62

78.33

400 features

68.09

77.89

0.4611

73.72

78.17

500 features

68.09

84.21

0.5326

77.34

78.33

All features

63.12

78.19

0.4182

71.73

75.16

  1. MCC - Matthew's correlation coefficient, CV-Cross validation