Table 15 The average recall, precision and \(F_{1}\)-score of WBC 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 0.947 (0.01) 1.000 (0.00) 0.973 (0.00) 0.953 (0.05) 0.948 (0.06) 0.950 (0.05) 0.930 (0.07) 0.943 (0.06) 0.934 (0.05)
Max norm 0.939 (0.01) 1.000 (0.00) 0.969 (0.00) 0.930 (0.05) 0.942 (0.05) 0.935 (0.04) 0.930 (0.05) 0.942 (0.05) 0.935 (0.04)
One norm 0.948 (0.02) 1.000 (0.00) 0.973 (0.01) 0.958 (0.06) 0.956 (0.04) 0.956 (0.04) 0.944 (0.07) 0.945 (0.04) 0.943 (0.05)
Mix norm 0.921 (0.02) 1.000 (0.00) 0.959 (0.01) 0.920 (0.06) 0.942 (0.05) 0.929 (0.04) 0.930 (0.05) 0.946 (0.05) 0.937 (0.04)
MoG 0.999 (0.00) 1.000 (0.00) 1.000 (0.00) 0.902 (0.07) 0.962 (0.05) 0.929 (0.05) 0.864 (0.08) 0.970 (0.04) 0.912 (0.06)
PDE 0.920 (0.07) 1.000 (0.00) 0.957 (0.04) 0.920 (0.07) 0.962 (0.04) 0.939 (0.05) 0.920 (0.07) 0.962 (0.04) 0.939 (0.05)
1-NN 1.000 (0.00) 1.000 (0.00) 1.000 (0.00) 0.859 (0.07) 0.567 (0.07) 0.682 (0.07) 0.854 (0.07) 0.552 (0.05) 0.670 (0.05)
k-NN 0.941 (0.01) 1.000 (0.00) 0.970 (0.01) 0.944 (0.04) 0.904 (0.05) 0.923 (0.03) 0.939 (0.06) 0.877 (0.08) 0.905 (0.05)
SVDD 0.959 (0.01) 1.000 (0.00) 0.979 (0.00) 0.912 (0.06) 0.922 (0.06) 0.916 (0.05) 0.921 (0.05) 0.907 (0.07) 0.913 (0.06)
5%
Two norm 0.962 (0.01) 0.986 (0.00) 0.974 (0.00) 0.953 (0.05) 0.928 (0.07) 0.939 (0.05) 0.958 (0.04) 0.941 (0.05) 0.949 (0.04)
Max norm 0.939 (0.01) 0.990 (0.00) 0.964 (0.00) 0.930 (0.05) 0.937 (0.06) 0.933 (0.05) 0.930 (0.05) 0.935 (0.05) 0.932 (0.05)
One norm 0.932 (0.02) 0.996 (0.00) 0.963 (0.01) 0.916 (0.07) 0.970 (0.03) 0.941 (0.05) 0.920 (0.07) 0.970 (0.02) 0.944 (0.04)
Mix norm 0.915 (0.03) 0.990 (0.00) 0.951 (0.02) 0.902 (0.07) 0.951 (0.04) 0.925 (0.05) 0.912 (0.06) 0.953 (0.05) 0.931 (0.04)
MoG 0.996 (0.01) 0.994 (0.01) 0.995 (0.00) 0.892 (0.06) 0.967 (0.04) 0.927 (0.04) 0.878 (0.09) 0.961 (0.07) 0.914 (0.07)
PDE 0.920 (0.07) 0.947 (0.00) 0.932 (0.04) 0.920 (0.07) 0.961 (0.03) 0.939 (0.05) 0.925 (0.07) 0.961 (0.03) 0.942 (0.05)
1-NN 1.000 (0.00) 0.947 (0.00) 0.973 (0.00) 0.995 (0.01) 0.472 (0.01) 0.641 (0.01) 0.995 (0.01) 0.472 (0.01) 0.641 (0.01)
k-NN 0.888 (0.02) 0.988 (0.01) 0.935 (0.01) 0.920 (0.06) 0.837 (0.06) 0.876 (0.05) 0.901 (0.09) 0.833 (0.08) 0.862 (0.06)
SVDD 0.940 (0.01) 0.967 (0.01) 0.953 (0.01) 0.884 (0.06) 0.925 (0.06) 0.902 (0.05) 0.883 (0.05) 0.919 (0.05) 0.900 (0.04)
10%
Two norm 0.920 (0.05) 0.987 (0.01) 0.951 (0.02) 0.921 (0.06) 0.967 (0.04) 0.942 (0.04) 0.935 (0.05) 0.962 (0.04) 0.947 (0.04)
Max norm 0.936 (0.01) 0.978 (0.01) 0.957 (0.01) 0.930 (0.05) 0.932 (0.05) 0.930 (0.04) 0.935 (0.05) 0.933 (0.05) 0.933 (0.04)
One norm 0.936 (0.01) 0.990 (0.00) 0.962 (0.01) 0.929 (0.07) 0.965 (0.03) 0.946 (0.05) 0.929 (0.07) 0.966 (0.03) 0.946 (0.04)
Mix norm 0.923 (0.01) 0.982 (0.01) 0.952 (0.01) 0.916 (0.07) 0.937 (0.06) 0.924 (0.05) 0.916 (0.06) 0.932 (0.06) 0.923 (0.05)
MoG 0.998 (0.00) 0.999 (0.00) 0.999 (0.00) 0.882 (0.09) 0.985 (0.02) 0.928 (0.05) 0.859 (0.10) 0.971 (0.03) 0.907 (0.06)
PDE 0.925 (0.07) 0.899 (0.00) 0.910 (0.04) 0.925 (0.07) 0.961 (0.04) 0.941 (0.05) 0.930 (0.07) 0.962 (0.03) 0.944 (0.04)
1-NN 1.000 (0.00) 0.899 (0.00) 0.947 (0.00) 0.995 (0.01) 0.473 (0.01) 0.642 (0.01) 0.995 (0.01) 0.472 (0.01) 0.641 (0.01)
k-NN 0.826 (0.03) 0.962 (0.01) 0.889 (0.02) 0.817 (0.09) 0.724 (0.09) 0.762 (0.07) 0.807 (0.10) 0.707 (0.06) 0.749 (0.05)
SVDD 0.944 (0.01) 0.939 (0.01) 0.941 (0.01) 0.893 (0.06) 0.920 (0.05) 0.905 (0.05) 0.879 (0.05) 0.902 (0.05) 0.889 (0.04)
15%
Two norm 0.965 (0.01) 0.955 (0.01) 0.960 (0.00) 0.963 (0.04) 0.917 (0.06) 0.939 (0.05) 0.963 (0.04) 0.915 (0.07) 0.937 (0.05)
Max norm 0.936 (0.01) 0.967 (0.00) 0.951 (0.01) 0.930 (0.05) 0.932 (0.05) 0.930 (0.04) 0.935 (0.05) 0.933 (0.05) 0.933 (0.04)
One norm 0.937 (0.01) 0.991 (0.00) 0.963 (0.01) 0.925 (0.07) 0.974 (0.03) 0.948 (0.05) 0.925 (0.07) 0.971 (0.03) 0.946 (0.05)
Mix norm 0.932 (0.01) 0.974 (0.01) 0.953 (0.01) 0.921 (0.06) 0.940 (0.05) 0.929 (0.04) 0.935 (0.04) 0.945 (0.05) 0.939 (0.04)
MoG 0.937 (0.01) 0.966 (0.01) 0.951 (0.00) 0.921 (0.05) 0.935 (0.04) 0.927 (0.04) 0.916 (0.04) 0.926 (0.05) 0.920 (0.04)
PDE 0.939 (0.06) 0.971 (0.00) 0.954 (0.03) 0.939 (0.06) 0.944 (0.04) 0.941 (0.04) 0.939 (0.06) 0.941 (0.05) 0.939 (0.05)
1-NN 1.000 (0.00) 0.849 (0.00) 0.918 (0.00) 0.995 (0.01) 0.472 (0.01) 0.641 (0.01) 0.995 (0.01) 0.472 (0.01) 0.641 (0.01)
k-NN 0.763 (0.02) 0.914 (0.02) 0.832 (0.02) 0.746 (0.08) 0.618 (0.03) 0.675 (0.05) 0.727 (0.12) 0.608 (0.10) 0.659 (0.10)
SVDD 0.904 (0.01) 0.959 (0.01) 0.930 (0.01) 0.870 (0.06) 0.917 (0.03) 0.891 (0.04) 0.860 (0.07) 0.917 (0.05) 0.886 (0.04)
20%
Two norm 0.940 (0.01) 0.972 (0.01) 0.955 (0.01) 0.944 (0.05) 0.953 (0.04) 0.948 (0.04) 0.934 (0.07) 0.948 (0.04) 0.940 (0.05)
Max norm 0.936 (0.01) 0.958 (0.01) 0.947 (0.01) 0.930 (0.05) 0.937 (0.05) 0.932 (0.04) 0.930 (0.05) 0.937 (0.06) 0.932 (0.04)
One norm 0.943 (0.02) 0.975 (0.02) 0.958 (0.01) 0.944 (0.07) 0.961 (0.02) 0.951 (0.04) 0.943 (0.08) 0.949 (0.03) 0.945 (0.04)
Mix norm 0.898 (0.02) 0.975 (0.01) 0.935 (0.01) 0.920 (0.06) 0.948 (0.05) 0.933 (0.05) 0.883 (0.06) 0.952 (0.04) 0.914 (0.03)
MoG 0.996 (0.00) 0.973 (0.02) 0.984 (0.01) 0.911 (0.07) 0.958 (0.04) 0.932 (0.05) 0.878 (0.09) 0.938 (0.07) 0.902 (0.05)
PDE 0.925 (0.07) 0.798 (0.00) 0.855 (0.03) 0.925 (0.07) 0.960 (0.03) 0.941 (0.05) 0.930 (0.07) 0.961 (0.03) 0.944 (0.05)
1-NN 1.000 (0.00) 0.798 (0.00) 0.887 (0.00) 0.995 (0.01) 0.473 (0.01) 0.642 (0.01) 0.995 (0.01) 0.472 (0.01) 0.641 (0.01)
k-NN 0.981 (0.01) 0.801 (0.01) 0.882 (0.01) 1.000 (0.00) 0.480 (0.01) 0.649 (0.01) 0.991 (0.03) 0.476 (0.01) 0.643 (0.02)
SVDD 0.922 (0.01) 0.919 (0.01) 0.920 (0.01) 0.888 (0.06) 0.886 (0.05) 0.886 (0.04) 0.884 (0.06) 0.871 (0.06) 0.875 (0.04)
  1. The bolded value is the best within the group