Table 14 The average recall, precision and \(F_{1}\)-score of Liver 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 0.908 (0.09) 1.000 (0.00) 0.950 (0.05) 0.929 (0.08) 0.445 (0.05) 0.601 (0.05) 0.937 (0.06) 0.439 (0.04) 0.596 (0.04)
Max norm 0.984 (0.01) 1.000 (0.00) 0.992 (0.00) 0.979 (0.03) 0.425 (0.01) 0.593 (0.02) 0.972 (0.05) 0.425 (0.02) 0.591 (0.02)
One norm 0.877 (0.02) 1.000 (0.00) 0.934 (0.01) 0.837 (0.16) 0.47 (0.05) 0.597 (0.07) 0.823 (0.18) 0.464 (0.05) 0.587 (0.08)
Mix norm 0.984 (0.01) 1.000 (0.00) 0.992 (0.00) 0.986 (0.03) 0.427 (0.01) 0.596 (0.01) 0.986 (0.03) 0.427 (0.01) 0.596 (0.01)
MoG 0.938 (0.01) 1.000 (0.00) 0.968 (0.01) 0.867 (0.12) 0.448 (0.04) 0.589 (0.05) 0.860 (0.12) 0.448 (0.04) 0.588 (0.06)
PDE 0.810 (0.16) 1.000 (0.00) 0.885 (0.11) 0.810 (0.16) 0.435 (0.04) 0.563 (0.07) 0.802 (0.16) 0.432 (0.04) 0.559 (0.07)
1-NN 1.000 (0.00) 1.000 (0.00) 1.000 (0.00) 0.958 (0.05) 0.428 (0.03) 0.591 (0.03) 0.951 (0.04) 0.428 (0.02) 0.590 (0.02)
k-NN 0.998 (0.00) 1.000 (0.00) 0.999 (0.00) 0.993 (0.02) 0.429 (0.02) 0.599 (0.02) 0.993 (0.02) 0.428 (0.02) 0.598 (0.02)
SVDD 0.427 (0.01) 1.000 (0.00) 0.598 (0.01) 0.274 (0.16) 0.534 (0.14) 0.346 (0.17) 0.289 (0.17) 0.557 (0.19) 0.364 (0.19)
5%
Two norm 0.993 (0.00) 0.947 (0.01) 0.970 (0.00) 0.993 (0.02) 0.428 (0.01) 0.598 (0.02) 0.986 (0.03) 0.421 (0.01) 0.590 (0.01)
Max norm 0.986 (0.00) 0.946 (0.00) 0.965 (0.00) 0.986 (0.03) 0.426 (0.01) 0.594 (0.02) 0.986 (0.03) 0.425 (0.01) 0.593 (0.02)
One norm 0.998 (0.00) 0.951 (0.00) 0.974 (0.00) 0.986 (0.03) 0.430 (0.01) 0.599 (0.01) 0.972 (0.05) 0.425 (0.02) 0.591 (0.02)
Mix norm 0.993 (0.01) 0.946 (0.00) 0.969 (0.00) 1.000 (0.00) 0.424 (0.01) 0.596 (0.01) 0.986 (0.03) 0.422 (0.01) 0.591 (0.02)
MoG 0.905 (0.01) 0.955 (0.01) 0.929 (0.01) 0.887 (0.13) 0.440 (0.04) 0.586 (0.06) 0.866 (0.13) 0.426 (0.04) 0.570 (0.06)
PDE 0.809 (0.18) 0.947 (0.01) 0.861 (0.12) 0.809 (0.18) 0.435 (0.05) 0.563 (0.08) 0.809 (0.16) 0.439 (0.04) 0.566 (0.07)
1-NN 1.000 (0.00) 0.947 (0.01) 0.973 (0.00) 0.979 (0.03) 0.423 (0.02) 0.590 (0.02) 0.972 (0.03) 0.422 (0.02) 0.589 (0.02)
k-NN 0.998 (0.00) 0.951 (0.01) 0.974 (0.00) 0.993 (0.02) 0.429 (0.02) 0.599 (0.02) 0.993 (0.02) 0.425 (0.02) 0.595 (0.02)
SVDD 0.628 (0.03) 0.956 (0.02) 0.757 (0.02) 0.281 (0.14) 0.535 (0.18) 0.340 (0.13) 0.296 (0.15) 0.571 (0.21) 0.364 (0.15)
10%
Two norm 0.986 (0.00) 0.900 (0.00) 0.941 (0.00) 0.986 (0.03) 0.428 (0.01) 0.597 (0.01) 0.986 (0.03) 0.428 (0.01) 0.597 (0.01)
Max norm 0.983 (0.01) 0.903 (0.01) 0.941 (0.00) 0.979 (0.03) 0.425 (0.01) 0.593 (0.01) 0.972 (0.05) 0.423 (0.02) 0.589 (0.02)
One norm 1.000 (0.00) 0.900 (0.00) 0.948 (0.00) 1.000 (0.00) 0.428 (0.01) 0.600 (0.01) 0.986 (0.03) 0.418 (0.01) 0.587 (0.01)
Mix norm 0.994 (0.01) 0.904 (0.01) 0.947 (0.01) 0.964 (0.06) 0.434 (0.02) 0.598 (0.03) 0.944 (0.07) 0.414 (0.02) 0.575 (0.03)
MoG 0.938 (0.01) 0.916 (0.01) 0.927 (0.01) 0.867 (0.12) 0.445 (0.02) 0.586 (0.04) 0.860 (0.12) 0.441 (0.03) 0.582 (0.05)
PDE 0.746 (0.16) 0.899 (0.01) 0.806 (0.10) 0.746 (0.16) 0.464 (0.03) 0.565 (0.06) 0.753 (0.16) 0.471 (0.03) 0.572 (0.06)
1-NN 1.000 (0.00) 0.899 (0.01) 0.947 (0.00) 0.972 (0.03) 0.425 (0.02) 0.591 (0.02) 0.965 (0.05) 0.422 (0.02) 0.587 (0.03)
k-NN 0.996 (0.00) 0.904 (0.01) 0.948 (0.00) 0.993 (0.02) 0.429 (0.01) 0.599 (0.02) 0.979 (0.05) 0.426 (0.01) 0.594 (0.02)
SVDD 0.633 (0.02) 0.910 (0.02) 0.747 (0.02) 0.281 (0.14) 0.536 (0.20) 0.341 (0.13) 0.296 (0.15) 0.526 (0.15) 0.363 (0.15)
15%
Two norm 0.968 (0.02) 0.846 (0.00) 0.903 (0.01) 0.979 (0.05) 0.432 (0.02) 0.599 (0.02) 0.965 (0.05) 0.424 (0.01) 0.589 (0.02)
Max norm 0.997 (0.00) 0.848 (0.01) 0.917 (0.00) 0.986 (0.03) 0.427 (0.01) 0.596 (0.01) 0.965 (0.07) 0.415 (0.02) 0.580 (0.03)
One norm 0.934 (0.07) 0.842 (0.01) 0.884 (0.04) 0.957 (0.06) 0.442 (0.05) 0.603 (0.05) 0.958 (0.05) 0.431 (0.02) 0.594 (0.03)
Mix norm 0.986 (0.00) 0.843 (0.00) 0.909 (0.00) 0.986 (0.03) 0.425 (0.01) 0.594 (0.01) 0.986 (0.03) 0.425 (0.01) 0.594 (0.01)
MoG 0.938 (0.02) 0.863 (0.01) 0.899 (0.01) 0.874 (0.11) 0.440 (0.03) 0.585 (0.05) 0.860 (0.12) 0.440 (0.04) 0.582 (0.06)
PDE 0.767 (0.17) 0.845 (0.01) 0.793 (0.10) 0.767 (0.17) 0.449 (0.06) 0.562 (0.09) 0.774 (0.18) 0.446 (0.06) 0.562 (0.09)
1-NN 1.000 (0.00) 0.845 (0.01) 0.916 (0.00) 0.986 (0.03) 0.430 (0.02) 0.599 (0.02) 0.986 (0.03) 0.427 (0.02) 0.596 (0.02)
k-NN 0.996 (0.00) 0.847 (0.00) 0.915 (0.00) 1.000 (0.00) 0.427 (0.01) 0.598 (0.01) 0.979 (0.05) 0.420 (0.01) 0.587 (0.01)
SVDD 0.681 (0.04) 0.860 (0.03) 0.759 (0.03) 0.296 (0.15) 0.504 (0.18) 0.362 (0.17) 0.303 (0.17) 0.535 (0.20) 0.375 (0.18)
20%
Two norm 0.966 (0.03) 0.798 (0.00) 0.874 (0.01) 0.979 (0.05) 0.430 (0.01) 0.597 (0.02) 0.972 (0.05) 0.426 (0.01) 0.592 (0.02)
Max norm 0.986 (0.00) 0.796 (0.00) 0.880 (0.00) 0.986 (0.03) 0.425 (0.01) 0.594 (0.02) 0.986 (0.03) 0.424 (0.02) 0.593 (0.02)
One norm 0.922 (0.11) 0.794 (0.01) 0.850 (0.06) 0.957 (0.06) 0.456 (0.08) 0.613 (0.07) 0.944 (0.07) 0.438 (0.04) 0.597 (0.04)
Mix norm 0.986 (0.00) 0.796 (0.00) 0.880 (0.00) 0.986 (0.03) 0.425 (0.01) 0.594 (0.02) 0.986 (0.03) 0.424 (0.02) 0.593 (0.02)
MoG 0.905 (0.01) 0.800 (0.01) 0.849 (0.01) 0.887 (0.13) 0.434 (0.04) 0.582 (0.07) 0.866 (0.13) 0.430 (0.04) 0.573 (0.06)
PDE 0.774 (0.18) 0.798 (0.00) 0.774 (0.10) 0.774 (0.18) 0.448 (0.05) 0.563 (0.09) 0.774 (0.18) 0.443 (0.06) 0.560 (0.09)
1-NN 1.000 (0.00) 0.798 (0.00) 0.888 (0.00) 0.993 (0.02) 0.420 (0.01) 0.590 (0.01) 1.000 (0.00) 0.423 (0.01) 0.594 (0.01)
k-NN 0.960 (0.01) 0.796 (0.01) 0.870 (0.01) 0.971 (0.07) 0.438 (0.02) 0.604 (0.03) 0.936 (0.08) 0.426 (0.03) 0.585 (0.04)
SVDD 0.625 (0.02) 0.805 (0.02) 0.704 (0.02) 0.317 (0.15) 0.467 (0.15) 0.368 (0.15) 0.282 (0.13) 0.480 (0.18) 0.343 (0.15)
  1. The bolded value is the best within the group