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

Feizi, Tayyebe; Moattar, Mohammad Hossein; Tabatabaee, Hamid

doi:10.1186/s40537-023-00832-2

Journal of Big Data

Table 10 The average F measure of the 3NN classifier in the multi-manifold and single manifold approaches

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

Dataset	PCA	LPP	NPE	Multi-Manifold
ecoli1	0.84 ± 0.160 (4)	0.85 ± 0.152 (1)	0.85 ± 0.152 (1)	0.85 ± 0.152 (1)
ecoli2	0.88 ± 0.132 (4)	0.90 ± 0.104 (2)	0.90 ± 0.104 (2)	0.91 ± 0.105 (1)
ecoli3	0.73 ± 0.223 (2)	0.71 ± 0.191 (3)	0.71 ± 0.191 (3)	0.86 ± 0.195 (1)
ecoli4	0.86 ± 0.296 (4)	0.89 ± 0.272 (1)	0.89 ± 0.272 (1)	0.92 ± 0.170 (1)
ecoli0147vs56	0.82 ± 0.150 (3)	0.82 ± 0.150 (3)	0.84 ± 0.159 (1)	0.89 ± 0.119 (1)
ecoli034_5	0.86 ± 0.180 (1)	0.86 ± 0.180 (1)	0.86 ± 0.180 (1)	0.88 ± 0.151 (1)
ecoli0147_2356	0.81 ± 0.159 (1)	0.81 ± 0.159 (1)	0.81 ± 0.159 (1)	0.81 ± 0.159 (1)
glass0	0.78 ± 0.077 (3)	0.79 ± 0.080 (1)	0.78 ± 0.076 (3)	0.79 ± 0.116 (1)
glass0123456	0.92 ± 0.153 (1)	0.92 ± 0.153 (1)	0.92 ± 0.153 (1)	0.92 ± 0.153 (1)
kddcup-buffer_overflow_vs_back	1 ± 0 (1)	1 ± 0 (1)	1 ± 0 (1)	1 ± 0 (1)
new-thyroid1	0.93 ± 0.160 (2)	0.93 ± 0.160 (2)	0.93 ± 0.160 (2)	0.98 ± 0.075 (1)
page-blocks-1-3_vs_4	0.87 ± 0.210 (2)	0.87 ± 0.210 (2)	0.87 ± 0.210 (2)	096 ± 0.080 (1)
pima	0.59 ± 0.113 (2)	0.58 ± 0.094 (3)	0.58 ± 0.092 (3)	0.65 ± 0.067 (1)
segment0	0.97 ± 0.023 (3)	0.98 ± 0.023 (1)	0.98 ± 0.023 (1)	0.98 ± 0.017 (1)
shuttle_2_vs_5	0.97 ± 0.100 (2)	0.97 ± 0.100 (2)	0.97 ± 0.100 (2)	1 ± 0 (1)
vehicle2–1	0.92 ± 0.050 (2)	0.92 ± 0.052 (2)	0.92 ± 0.052 (2)	0.97 ± 0.030 (1)
vowel0	0.85 ± 0.174 (4)	0.87 ± 0.157 (2)	0.87 ± 0.157 (2)	0.90 ± 0.104 (1)
wisconsin	0.97 ± 0.050 (1)	0.97 ± 0.050 (1)	0.97 ± 0.050 (1)	0.97 ± 0.050 (1)
Average Ratings	2.00 (4)	1.33 (2)	1.28 (2)	1 (1)
Average F-measure	0.87 (2)	0.87 (2)	0.87 (2)	0.90 (1)

Back to article page