Research on imbalance machine learning methods for MR $$T_1$$ WI soft tissue sarcoma data

BMC Medical Imaging

Table 7 Running time of different machine learning models

Number	Model	Conventional Split-Running time (s)	SRS-Running time (s)
1	RFE+ROSE+ERT	64	65
2	RFE+SMOTE+ERT	66	66
3	RFE+STT+ERT	65	66
4	RFE+ADASYN+ERT	67	68
5	RFE+ROSE+RF	67	67
6	RFE+SMOTE+RF	69	67
7	RFE+STT+RF	66	68
8	RFE+ADASYN+RF	67	67
9	RFE+ROSE+BRF	66	66
10	RFE+SMOTE+BRF	66	66
11	RFE+STT+BRF	66	66
12	RFE+ADASYN+BRF	67	70
13	RFE+ROSE+SVM	68	66
14	RFE+SMOTE+SVM	64	66
15	RFE+STT+SVM	66	66
16	RFE+ADASYN+SVM	64	67
17	RFE+GDO+SVM	66	65

RFE recursive feature elimination; ROSE random oversampling examples; SMOTE synthetic minority oversampling technique; STT SMOTETomek; ADASYN adaptive synthetic samping; ERT extremely randomized trees; RF random forest; BRF balanced random forest; SVM support vector machine

ISSN: 1471-2342