Skip to main content

Table 2 Comparison of the performance of RF, SVM and ANN under all encoding procedures with both balanced and imbalanced training dataset

From: Prediction of donor splice sites using random forest with a new sequence encoding approach

EP MLA Balanced Dataset Imbalanced Dataset
TPR TNR F (α = 1) F (β = 2) G-mean WA MCC TPR TNR F (α = 1) F (β = 2) G-mean WA MCC
P-1 RF 0.954 0.924 0.940 0.932 0.939 0.939 0.878 0.842 0.896 0.865 0.880 0.869 0.869 0.739
(0.014) (0.014) (0.010) (0.012) (0.010) (0.010) (0.020) (0.064) (0.018) (0.032) (0.049) (0.030) (0.028) (0.043)
SVM 0.935 0.930 0.933 0.931 0.933 0.933 0.865 0.104 0.982 0.185 0.349 0.320 0.543 0.180
(0.015) (0.017) (0.015) (0.015) (0.016) (0.016) (0.031) (0.027) (0.018) (0.041) (0.031) (0.040) (0.013) (0.061)
ANN 0.892 0.896 0.894 0.895 0.894 0.894 0.787 0.032 0.988 0.061 0.136 0.178 0.510 0.068
(0.064) (0.080) (0.063) (0.062) (0.066) (0.065) (0.129) (0.026) (0.010) (0.046) (0.032) (0.065) (0.011) (0.055)
P-2 RF 0.937 0.901 0.920 0.911 0.919 0.919 0.838 0.883 0.894 0.888 0.891 0.888 0.889 0.777
(0.020) (0.016) (0.016) (0.018) (0.016) (0.016) (0.033) (0.038) () (0.025) (0.030) (0.019) (0.019) (0.035)
SVM 0.720 0.773 0.740 0.752 0.746 0.746 0.493 0.321 0.989 0.482 0.689 0.563 0.655 0.417
(0.029) (0.106) (0.041) (0.026) (0.049) (0.051) (0.108) (0.051) (0.008) (0.055) (0.053) (0.043) (0.025) (0.048)
ANN 0.775 0.777 0.776 0.776 0.776 0.776 0.552 0.305 0.978 0.460 0.661 0.546 0.642 0.383
(0.067) (0.037) (0.049) (0.059) (0.048) (0.045) (0.090) (0.049) (0.014) (0.052) (0.051) (0.043) (0.022) (0.046)
P-3 RF 0.940 0.908 0.925 0.917 0.924 0.924 0.848 0.879 0.891 0.884 0.888 0.885 0.885 0.770
(0.017) (0.015) (0.012) (0.014) (0.012) (0.012) (0.246) (0.044) (0.022) (0.029) (0.034) (0.022) (0.022) (0.042)
SVM 0.789 0.807 0.796 0.800 0.798 0.798 0.595 0.249 0.988 0.395 0.609 0.496 0.619 0.352
(0.044) (0.068) (0.042) (0.042) (0.046) (0.045) (0.090) (0.052) (0.008) (0.062) (0.056) (0.049) (0.026) (0.055)
ANN 0.757 0.760 0.758 0.759 0.759 0.759 0.517 0.272 0.979 0.421 0.626 0.516 0.626 0.355
(0.118) (0.099) (0.067) (0.098) (0.057) (0.048) (0.086) (0.066) (0.009) (0.081) (0.072) (0.064) (0.034) (0.076)
  1. The values inside the brackets () are the standard errors
  2. EP encoding procedure, MLA machine learning approaches