Table 12 The average recall, precision and \(F_{1}\)-score of Iris-versicolor data depending on the noise level (standard deviation in parentheses)

From: Norm ball classifier for one-class classification

Classifier Training Validation Test
Recall Precision \({F}_{1}\)-score Recall Precision \({F}_{1}\)-score Recall Precision \({F}_{1}\)-score
0%
Two norm 0.987 (0.02) 1.000 (0.00) 0.993 (0.01) 0.900 (0.09) 0.934 (0.09) 0.909 (0.05) 0.880 (0.12) 0.828 (0.17) 0.837 (0.10)
Max norm 0.993 (0.01) 1.000 (0.00) 0.997 (0.01) 0.880 (0.08) 0.905 (0.12) 0.883 (0.05) 0.840 (0.08) 0.859 (0.15) 0.845 (0.11)
One norm 0.973 (0.02) 1.000 (0.00) 0.986 (0.01) 0.980 (0.04) 0.917 (0.08) 0.944 (0.04) 0.900 (0.11) 0.851 (0.12) 0.867 (0.09)
Mix norm 0.993 (0.01) 1.000 (0.00) 0.997 (0.01) 1.000 (0.00) 0.781 (0.13) 0.871 (0.09) 0.960 (0.05) 0.752 (0.21) 0.830 (0.16)
MoG 0.993 (0.01) 1.000 (0.00) 0.997 (0.01) 0.980 (0.04) 0.925 (0.11) 0.947 (0.06) 0.960 (0.05) 0.863 (0.13) 0.903 (0.08)
PDE 1.000 (0.00) 1.000 (0.00) 1.000 (0.00) 1.000 (0.00) 0.902 (0.09) 0.946 (0.05) 0.980 (0.04) 0.854 (0.15) 0.907 (0.10)
1-NN 1.000 (0.00) 1.000 (0.00) 1.000 (0.00) 0.860 (0.10) 0.830 (0.13) 0.830 (0.04) 0.900 (0.15) 0.772 (0.20) 0.796 (0.11)
k-NN 0.973 (0.02) 1.000 (0.00) 0.986 (0.01) 0.960 (0.05) 0.838 (0.13) 0.887 (0.06) 0.960 (0.05) 0.799 (0.17) 0.863 (0.11)
SVDD 0.940 (0.04) 1.000 (0.00) 0.969 (0.02) 0.900 (0.11) 0.892 (0.10) 0.891 (0.08) 0.860 (0.10) 0.900 (0.13) 0.871 (0.08)
5%
Two norm 0.980 (0.02) 1.000 (0.00) 0.990 (0.01) 0.960 (0.08) 0.851 (0.18) 0.886 (0.10) 0.960 (0.08) 0.838 (0.16) 0.883 (0.10)
Max norm 0.947 (0.04) 0.980 (0.04) 0.963 (0.04) 0.940 (0.08) 0.831 (0.17) 0.874 (0.12) 0.900 (0.11) 0.860 (0.17) 0.873 (0.14)
One norm 1.000 (0.00) 1.000 (0.00) 1.000 (0.00) 1.000 (0.00) 0.906 (0.10) 0.948 (0.06) 0.980 (0.04) 0.879 (0.15) 0.919 (0.10)
Mix norm 0.947 (0.05) 0.979 (0.04) 0.962 (0.04) 0.900 (0.06) 0.893 (0.12) 0.895 (0.09) 0.820 (0.12) 0.800 (0.23) 0.791 (0.16)
MoG 1.000 (0.00) 1.000 (0.00) 1.000 (0.00) 0.980 (0.04) 0.892 (0.15) 0.925 (0.09) 1.000 (0.00) 0.818 (0.17) 0.890 (0.10)
PDE 0.920 (0.07) 1.000 (0.00) 0.957 (0.04) 0.920 (0.07) 0.876 (0.16) 0.893 (0.12) 0.920 (0.07) 0.861 (0.16) 0.885 (0.11)
1-NN 1.000 (0.00) 0.944 (0.02) 0.971 (0.01) 0.860 (0.10) 0.472 (0.20) 0.593 (0.18) 0.900 (0.15) 0.392 (0.10) 0.544 (0.12)
k-NN 0.973 (0.02) 1.000 (0.00) 0.986 (0.01) 0.960 (0.05) 0.845 (0.17) 0.886 (0.10) 0.960 (0.05) 0.801 (0.20) 0.862 (0.14)
SVDD 0.853 (0.02) 1.000 (0.00) 0.921 (0.01) 0.860 (0.08) 0.954 (0.09) 0.897 (0.03) 0.820 (0.15) 0.967 (0.07) 0.875 (0.08)
10%
Two norm 0.967 (0.02) 1.000 (0.00) 0.983 (0.01) 0.960 (0.08) 0.818 (0.15) 0.873 (0.09) 0.960 (0.08) 0.818 (0.15) 0.873 (0.09)
Max norm 0.967 (0.04) 1.000 (0.00) 0.983 (0.02) 0.900 (0.11) 0.853 (0.18) 0.858 (0.10) 0.920 (0.10) 0.836 (0.16) 0.868 (0.11)
One norm 0.920 (0.02) 1.000 (0.00) 0.958 (0.01) 0.900 (0.11) 0.867 (0.14) 0.870 (0.08) 0.840 (0.08) 0.888 (0.12) 0.860 (0.09)
Mix norm 0.833 (0.10) 1.000 (0.00) 0.906 (0.06) 0.800 (0.14) 0.933 (0.09) 0.855 (0.10) 0.64 (0.27) 0.894 (0.21) 0.683 (0.22)
MoG 0.993 (0.01) 1.000 (0.00) 0.997 (0.01) 0.980 (0.04) 0.900 (0.13) 0.931 (0.07) 0.960 (0.05) 0.849 (0.14) 0.895 (0.08)
PDE 0.920 (0.07) 1.000 (0.00) 0.957 (0.04) 0.920 (0.07) 0.885 (0.14) 0.899 (0.11) 0.900 (0.09) 0.859 (0.16) 0.873 (0.12)
1-NN 1.000 (0.00) 0.893 (0.01) 0.943 (0.01) 0.860 (0.10) 0.469 (0.17) 0.595 (0.16) 0.900 (0.15) 0.405 (0.08) 0.558 (0.10)
k-NN 0.973 (0.02) 1.000 (0.00) 0.986 (0.01) 0.960 (0.05) 0.833 (0.14) 0.882 (0.08) 0.960 (0.05) 0.760 (0.16) 0.841 (0.11)
SVDD 0.773 (0.01) 1.000 (0.00) 0.872 (0.01) 0.800 (0.13) 1.000 (0.00) 0.883 (0.08) 0.760 (0.15) 0.982 (0.04) 0.846 (0.08)
15%
Two norm 0.927 (0.02) 0.993 (0.01) 0.959 (0.01) 0.920 (0.08) 0.833 (0.21) 0.862 (0.14) 0.880 (0.08) 0.792 (0.22) 0.813 (0.12)
Max norm 0.907 (0.08) 0.954 (0.04) 0.927 (0.04) 0.920 (0.08) 0.832 (0.18) 0.860 (0.11) 0.860 (0.14) 0.788 (0.18) 0.809 (0.12)
One norm 0.907 (0.06) 0.964 (0.05) 0.934 (0.05) 0.860 (0.08) 0.960 (0.08) 0.905 (0.07) 0.760 (0.08) 0.879 (0.15) 0.812 (0.10)
Mix norm 0.907 (0.06) 0.974 (0.04) 0.937 (0.04) 0.880 (0.08) 0.836 (0.11) 0.856 (0.08) 0.780 (0.12) 0.789 (0.27) 0.757 (0.17)
MoG 1.000 (0.00) 1.000 (0.00) 1.000 (0.00) 0.940 (0.08) 0.954 (0.09) 0.941 (0.05) 0.900 (0.09) 0.881 (0.15) 0.879 (0.07)
PDE 0.920 (0.07) 1.000 (0.00) 0.957 (0.04) 0.920 (0.07) 0.877 (0.16) 0.892 (0.11) 0.900 (0.09) 0.856 (0.18) 0.866 (0.11)
1-NN 1.000 (0.00) 0.848 (0.01) 0.918 (0.01) 1.000 (0.00) 0.338 (0.01) 0.505 (0.01) 1.000 (0.00) 0.336 (0.00) 0.503 (0.01)
k-NN 0.860 (0.08) 1.000 (0.00) 0.923 (0.05) 0.900 (0.06) 0.872 (0.16) 0.873 (0.07) 0.800 (0.11) 0.823 (0.18) 0.791 (0.07)
SVDD 0.827 (0.02) 1.000 (0.00) 0.905 (0.01) 0.840 (0.15) 0.954 (0.09) 0.879 (0.08) 0.840 (0.15) 0.915 (0.13) 0.856 (0.07)
20%
Two norm 0.927 (0.02) 0.973 (0.02) 0.949 (0.01) 0.900 (0.06) 0.825 (0.18) 0.854 (0.13) 0.880 (0.08) 0.737 (0.24) 0.777 (0.15)
Max norm 0.953 (0.05) 0.902 (0.06) 0.926 (0.05) 0.940 (0.08) 0.815 (0.14) 0.864 (0.09) 0.920 (0.12) 0.715 (0.14) 0.799 (0.12)
One norm 0.927 (0.05) 0.954 (0.06) 0.939 (0.05) 0.820 (0.08) 0.956 (0.05) 0.881 (0.06) 0.760 (0.15) 0.836 (0.19) 0.773 (0.13)
Mix norm 0.887 (0.13) 0.934 (0.09) 0.908 (0.11) 0.900 (0.11) 0.837 (0.17) 0.849 (0.11) 0.780 (0.22) 0.793 (0.22) 0.756 (0.20)
MoG 1.000 (0.00) 1.000 (0.00) 1.000 (0.00) 0.940 (0.08) 0.948 (0.07) 0.940 (0.04) 0.900 (0.09) 0.875 (0.12) 0.876 (0.05)
PDE 0.880 (0.07) 1.000 (0.00) 0.935 (0.04) 0.880 (0.07) 0.861 (0.13) 0.867 (0.10) 0.900 (0.09) 0.844 (0.15) 0.864 (0.11)
1-NN 1.000 (0.00) 0.794 (0.02) 0.885 (0.01) 0.760 (0.19) 0.412 (0.18) 0.525 (0.18) 0.760 (0.22) 0.350 (0.12) 0.472 (0.13)
k-NN 0.907 (0.02) 1.000 (0.00) 0.951 (0.01) 0.940 (0.05) 0.818 (0.15) 0.865 (0.09) 0.920 (0.07) 0.802 (0.17) 0.847 (0.11)
SVDD 0.880 (0.03) 1.000 (0.00) 0.936 (0.02) 0.920 (0.12) 0.936 (0.09) 0.919 (0.06) 0.840 (0.08) 0.885 (0.10) 0.858 (0.07)
  1. The bolded value is the best within the group