Skip to main content

Table 1 Performance of the proposed models

From: Deep neural networks for human microRNA precursor detection

Model (Data set partition)

Sen.(%)

Spe.(%)

F1(%)

MCC(%)

Acc.(%)

CNN (Training/Evalu./Test)

88.83

88.28

88.83

77.11

88.56

CNN (10-fold CV)

89.58 ± 4.72

84.90 ± 4.84

87.53 ± 1.38

74.72 ± 3.52

87.24 ± 1.80

RNN (Training/Evalu./Test)

85.71

91.28

88.35

77.03

88.43

RNN (10-fold CV)

85.89 ± 3.29

91.14 ± 2.75

88.09 ± 2.03

77.04 ± 3.66

88.44 ± 1.80

  1. Note: Classification performance of different models on the testing dataset was shown as sensitivity (column 2), specificity (column 3), F1-Score (column 4), MCC (column 5) and accuracy (column 6) respectively. For the 10-fold CV, the performance was shown as mean ± standard error