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

Table 4

with different oversampling methods from real datasets.

Id	NO-RS	SMOTE	BLSMOTE	ADASYN	DBSMOTE	AWSMOTE
1	0.9691	0.9742	0.9617	0.9622	0.9740	0.9748
2	0.8820	0.9056	0.8861	0.9039	0.9130	0.9156
3	0.9559	0.9903	——	0.9952	0.9950	0.9972
4	0.9676	0.9641	——	0.9599	0.9700	0.9711
5	0.9065	0.9146	0.9393	0.9446	0.9250	0.9503
6	0.9117	0.8789	0.9181	0.9230	0.9137	0.9194
7	0.9503	0.9927	0.9859	0.9930	0.9881	0.9933
8	0.8339	0.8516	0.8962	0.9010	0.8823	0.9082
9	0.9143	0.9333	0.9391	0.9287	0.9110	0.9434
10	0.9765	0.9573	0.9724	0.9724	0.9745	0.9769
11	0.9478	0.9545	0.9543	0.9525	0.9531	0.9559
12	0.9769	0.9929	——	0.9832	0.9953	0.9944
13	0.9655	0.9870	——	0.9857	0.9871	0.9873
14	0.9761	0.9962	——	0.9951	0.9903	0.9977
15	0.9622	0.9769	——	0.9995	0.9987	1
16	0.9566	0.9939	——	0.9936	0.9939	0.9942
17	0.9537	0.9987	——	0.9985	0.9902	1
18	0.9898	0.9878	——	0.9860	0.9561	0.9935
19	0.9870	0.9797	——	0.9903	0.9886	0.9941
20	0.9756	——	——	——	0.9925	0.9960
21	0.9863	0.9800	——	0.9777	0.9908	0.9874
22	0.9705	0.9930	——	0.9941	0.9941	0.9946