Table A1.

Mean of the relative differences of the DOB-SCV and SCV per oversampler on the KEEL data sets.

	DTree				kNN
Sampler	% ${\mathit{diff}}_{\mathit{AUC}}$	% ${\mathit{diff}}_{\mathit{Acc}}$	% ${\mathit{diff}}_{\mathit{F}\mathit{1}}$	% ${\mathit{diff}}_{\mathit{G}}$	% ${\mathit{diff}}_{\mathit{AUC}}$	% ${\mathit{diff}}_{\mathit{Acc}}$	% ${\mathit{diff}}_{\mathit{F}\mathit{1}}$	% ${\mathit{diff}}_{\mathit{G}}$
Assembled-SMOTE	0.79	0.21	1.16	0.65	0.89	0.36	1.65	0.97
CCR	0.70	0.14	1.36	0.96	1.02	0.29	1.93	1.21
G-SMOTE	1.05	0.26	1.91	1.10	0.89	0.34	1.83	1.09
LVQ-SMOTE	0.57	0.27	1.49	0.58	0.89	0.39	2.27	1.32
Lee	0.98	0.26	1.71	0.87	0.87	0.38	1.61	0.97
ModularOverSampler	0.71	0.26	1.49	0.92	0.96	0.39	2.06	1.37
ProWSyn	0.93	0.23	1.27	0.95	0.86	0.43	1.73	1.00
Polynomial fitting SMOTE	0.80	0.14	1.47	1.11	0.95	0.31	2.09	1.31
SMOBD	0.86	0.23	1.44	0.82	0.85	0.35	1.47	0.89
SMOTE	0.96	0.28	1.69	0.79	0.86	0.34	1.57	0.96
SMOTE-IPF	0.92	0.24	1.21	0.86	0.85	0.37	1.54	0.90
SMOTE-TomekLinks	0.91	0.19	1.21	0.94	0.90	0.35	1.61	0.99
	SVM				MLP
Assembled-SMOTE	0.48	0.29	1.70	0.54	0.52	0.18	1.16	0.53
CCR	0.64	0.26	1.19	0.56	0.65	0.18	1.00	0.64
G-SMOTE	0.48	0.24	1.73	0.73	0.51	0.18	1.15	0.68
LVQ-SMOTE	0.46	0.12	1.04	0.48	0.50	0.21	1.00	0.56
Lee	0.46	0.18	1.61	0.65	0.51	0.19	1.19	0.55
ModularOverSampler	0.41	0.25	1.70	0.63	0.59	0.27	1.42	0.69
ProWSyn	0.45	0.24	1.49	0.54	0.55	0.25	1.36	0.68
Polynomial fitting SMOTE	0.45	0.14	1.44	0.50	0.58	0.27	1.46	0.64
SMOBD	0.43	0.11	1.49	0.48	0.49	0.23	1.34	0.68
SMOTE	0.38	0.28	1.54	0.49	0.83	0.37	1.54	0.78
SMOTE-IPF	0.44	0.32	1.50	0.42	0.61	0.29	1.24	0.57
SMOTE-TomekLinks	0.45	−0.06	1.69	0.53	0.59	0.23	1.23	0.79