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

Table 5

Time taken by different feature extraction methods.

Dataset	Feature extraction method	Feature size	Time
			CWRU	LBP	59	—	87.6
NMF + NLE	28	16.469		39525.0
LBP + NMF + NLE (FDL)	28	0.037		88.8
NMFSC + NLE [27]	25	8.089		19413.0
MFPT	LBP	59	—	93.6
	NMF + NLE	22	3.897	37413.6
	LBP + NMF + NLE (FDL)	22	0.012	95.4
	NMFSC + NLE [27]	100	2.247	21573.6