An Improved Bearing Fault Diagnosis Model of Variational Mode Decomposition Based on Linked Extension Neural Network

Table 8

Average scores of models on datasets in different imbalance degrees.

No.	Target	LENN	K-means	FCM	DBSCAN
Dataset 1	RI	0.9485	0.5191	0.5301	0.8278
	Macro-R	0.9700	0.4960	0.4960	0.9140
	Macro-F1	0.9667	0.4933	0.4933	0.9333
Dataset 2	RI	0.9314	0.4806	0.4814	0.8113
	Macro-R	0.9510	0.4950	0.4950	0.8901
	Macro-F1	0.9487	0.4827	0.4827	0.9052
Dataset 3	RI	0.9250	0.4522	0.4376	0.6839
	Macro-R	0.9322	0.4950	0.4950	0.8130
	Macro-F1	0.9367	0.4827	0.4827	0.8486
Dataset 4	RI	0.8920	0.3822	0.2024	0.4329
	Macro-R	0.9100	0.4940	0.4860	0.7500
	Macro-F1	0.9357	0.4786	0.4500	0.8929