Table 1 Five-fold cross-validation and Independent evaluation (IE) test results of the SVM method for oogenesis 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

0.02

82.8

82.86

83.15

65.57

83.33

84.62

80.88

66.7

2

0.001

84.06

83.21

85.04

68.13

83.33

84.62

80.88

66.7

3

0.02

85.33

83.57

86.99

70.71

84.06

86.15

81.16

68.23

4

0.02

82.05

80.71

83.39

63.87

86.23

86.15

84.85

72.4

5

0.001

81.89

81.79

82.37

63.76

82.61

84.62

79.71

65.32

Average

0.01

84

83

85

79.4

84

86

82

67.87

  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