A multi-manifold learning based instance weighting and under-sampling for imbalanced data classification problems

Journal of Big Data

Table 8 The Average performance measures of the 3NN classifier with the multi-manifold approach with reduction step of 10 percent

Name	Recall	Precision	G-means	F-measure	Accuracy
ecoli1	0.98 ± 0.050	0.70 ± 0.208	0.89 ± 0.126	0.85 ± 0.152	0.86 ± 0.152
ecoli2	0.95 ± 0.110	0.75 ± 0.200	0.93 ± 0.076	0.89 ± 0.116	0.93 ± 0.066
ecoli3	0.93 ± 0.225	0.56 ± 0.359	0.88 ± 0.155	0.73 ± 0.250	0.87 ± 0.114
ecoli4	0.95 ± 0.150	0.74 ± 0.350	0.92 ± 0.178	0.88 ± 0.279	0.91 ± 0.199
ecoli0147vs56	0.88 ± 0.183	0.71 ± 0.258	0.92 ± 0.106	0.84 ± 0.159	0.95 ± 0.039
ecoli034_5	0.90 ± 0.200	0.78 ± 0.269	0.92 ± 0.122	0.86 ± 0.180	0.95 ± 0.067
ecoli0147_2356	0.82 ± 0.240	0.64 ± 0.261	0.87 ± 0.147	0.81 ± 0.159	0.93 ± 0.049
glass0	0.89 ± 0.124	0.61 ± 0.121	0.77 ± 0.093	0.78 ± 0.076	0.75 ± 0.101
glass0123456	0.96 ± 0.080	0.83 ± 0.226	0.93 ± 0.115	0.91 ± 0.151	0.92 ± 0.131
kddcup-buffer_overflow_vs_back	1 ± 0	1 ± 0	1 ± 0	1 ± 0	1 ± 0
new-thyroid1	1 ± 0	0.93 ± 0.155	0.99 ± 0.026	0.98 ± 0.075	0.98 ± 0.043
page-blocks-1-3_vs_4	0.93 ± 0.133	0.71 ± 0.319	0.94 ± 0.084	0.87 ± 0.210	0.95 ± 0.071
Pima	0.70 ± 0.162	0.57 ± 0.142	0.66 ± 0.095	0.63 ± 0.094	0.68 ± 0.110
segment0	0.99 ± 0.027	0.90 ± 0.066	0.98 ± 0.014	0.97 ± 0.023	0.98 ± 0.012
shuttle_2_vs_5	1 ± 0	0.91 ± 0.269	0.99 ± 0.021	0.97 ± 0.100	0.98 ± 0.040
vehicle2–1	0.98 ± 0.040	0.80 ± 0.110	0.94 ± 0.034	0.92 ± 0.048	0.92 ± 0.047
vowel0	0.97 ± 0.245	0.62 ± 0.301	0.84 ± 0.176	0.87 ± 0.212	0.81 ± 0.117
Wisconsin	0.98 ± 0.028	0.93 ± 0.079	0.94 ± 0.086	0.97 ± 0.039	0.94 ± 0.061
Average	0.933889	0.761667	0.906111	0.873889	0.906111