Table 2 Performance evaluation of XGBoost model using tenfold cross-validation.

From: A hybrid residue based sequential encoding mechanism with XGBoost improved ensemble model for identifying 5-hydroxymethylcytosine modifications

Method

ACC (%)

SN (%)

Precision (%)

SPE (%)

F1 Score (%)

MCC

AUC

Mismatch

75.13

65.21

78.93

82.46

71.34

0.573

0.813

ANF

77.34

68.34

76.83

84.75

73.05

0.612

0.837

PSTNPss

78.09

70.65

79.28

86.85

75.75

0.652

0.855

ASDC

82.94

75.03

82.48

88.85

80.92

0.724

0.884

DAC

85.95

80.72

85.58

90.57

83.32

0.785

0.912

Hybrid feature (before feature selection)

86.87

82.12

87.52

92.75

85.41

0.814

0.932

Hybrid feature (after feature selection)

89.97

87.78

90.94

94.45

89.34

0.876

0.953

Back to article page