Table 4 Five-fold cross-validation and Independent evaluation (IE) test results of the SVM method for general datasets.

From: PrESOgenesis: A two-layer multi-label predictor for identifying fertility-related proteins using support vector machine and pseudo amino acid composition approach

Datasets

λ*

Five-fold cross-validation test

Independent evaluation test

Accuracy (%)

Sensitivity (%)

Specificity (%)

MCC (%)

Accuracy (%)

Sensitivity (%)

Specificity (%)

MCC (%)

1

0.001

82.94

84.27

82.48

65.87

81.02

80.41

79.48

64.96

2

0.01

82.56

84.34

81.84

70.1

84.46

83.92

83.19

68.82

3

0.03

82.21

82.85

82.23

64.42

82.26

82.75

80.17

64.48

4

0.04

83.31

83.46

83.63

66.62

82.53

81.29

81.52

64.93

5

0.05

83.87

82.85

84.98

67.76

83.77

80.12

84.57

67.41

Average

0.03

82.97

83.55

83.03

66.95

82.88

81.69

81.78

66.12

  1. *The optimum λ parameter value of kernel function of SVM using a grid-search technique based on five-fold cross-validation. Also, the optimum parameter C value was obtained 100 in all of models.
Back to article page