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

Table 7

Precision with different oversampling methods from real datasets.

Id	NO-RS	SMOTE	BLSMOTE	ADASYN	DBSMOTE	AWSMOTE
1	0.9611	0.9740	0.9876	0.9701	0.9578	0.9676
2	0.8832	0.9684	0.9353	0.9800	0.9380	0.9813
3	0.9156	0.9974	——	0.9998	0.9942	1
4	0.9717	0.9724	——	1	0.9794	1
5	0.8402	0.9240	0.9440	0.9780	0.8860	0.9825
6	0.8941	0.9397	0.9553	0.9470	0.9029	0.9674
7	0.9097	0.9998	0.9779	1	0.9860	0.9989
8	0.7177	0.8393	0.8289	0.8769	0.8196	0.9015
9	0.8953	0.9618	0.9502	0.9434	0.8916	0.9735
10	0.9752	0.9459	0.9948	1	0.9900	1
11	0.9449	0.9561	0.9786	0.9556	0.9624	0.9648
12	0.9553	1	——	0.9954	0.9948	1
13	0.9333	0.9990	——	1	0.9932	1
14	0.9534	1	——	1	0.9899	1
15	0.9273	0.9554	——	1	0.9991	1
16	0.9169	1	——	1	0.9978	0.9980
17	0.9116	0.9987	——	0.9989	0.9810	1
18	0.9800	0.9943	——	0.9973	0.9448	0.9975
19	0.9745	0.9853	——	0.9997	0.9842	1
20	0.9523	——	——	——	1	1
21	0.9777	0.9914	——	1	0.9945	1
22	0.9428	1	——	1	0.9894	1