AWSMOTE: An SVM-Based Adaptive Weighted SMOTE for Class-Imbalance Learning

Table 3

Summary description of real datasets.

Id	Dataset	Total	Minority	Majority	Features	IR
1	WDBC	569	212	357	30	1.68
2	Libras movement	360	120	240	90	2
3	Accent	329	29	300	12	10.34
4	Seeds	210	70	140	7	2
5	Speech features	756	192	564	752	2.93
6	Musk1	476	207	269	166	1.29
7	Musk2	321	52	269	167	5.17
8	CHD	72	51	21	77	2.42
9	HNP	171	104	67	42	1.55
10	Vehicle0	846	199	647	18	3.52
11	Glass0123vs456	214	52	162	8	3.11
12	Newthyroid1	215	35	180	5	5.14
13	Glass4	214	13	201	9	15.46
14	Glass5	214	9	205	9	22.77
15	Shuttlec2vsc4	129	6	123	9	20.5
16	Cleveland0vs4	177	13	164	13	12.61
17	Glass06vs5	108	9	99	9	11
18	Ecoli0146vs5	260	20	240	6	13
19	Ecoli046vs5	203	20	183	7	9.15
20	Zoo3	101	5	96	16	19.2
21	Winewhite3vs7	900	20	880	11	44
22	Winewhite9vs4	168	6	162	11	27