Rolling Bearing Fault Diagnosis Method Based on Multisynchrosqueezing S Transform and Faster Dictionary Learning

Table 6

Comparison of the proposed method with other methods.

Dataset	Method	Sparsity	Feature size	Fault types	Accuracy (%)
CWRU	The proposed method	—	28	10	99
	SFC-DL + SVM [27]	0.7	25	10	98.03
	HHT + CNN [38]	—	32 × 32	10	95
	WPE-MF + SVM [40]	—	33	10	88.9
MFPT	The proposed method	—	22	3	97.85
	SFC-DL + SVM [27]	0.7	100	3	95.83
	HHT + CNN [39]	—	96 × 96	3	92.9
			32 × 32	3	75.9