Combining DBN and FCM for Fault Diagnosis of Roller Element Bearings without Using Data Labels

Table 5

The results of classification accuracy.

Mode	Dataset	The accuracy of each cluster (%)										Total (%)
		NR	BF1	IRF1	ORF2	BF2	IRF2	ORF2	BF3	IRF3	ORF3
EEMD-FE-PCA-FCM (k = 2)	A	93.3	100	60	100	100	100	93.3	100	100	26.7	87.33
	B	100	100	100	63.3	100	73.3	100	100	100	30	86.69
	C	100	100	56.7	43.3	96.77	96.7	66.7	100	100	100	76.34
MD-SVD-FCM (k = 2)	A	100	100	100	70	100	100	33.3%	100	23.3	13.3	77
	B	90	100	100	56.7	100	86.7	100	100	30	43.3	80.67
	C	76.7	100	100	76.7	23.67	100	23.3	100	60	100	76.037
DBN-PCA-FCM (k = 2)	A	100	100	96.7	100	100	100	100	100	90	100	98.67
	B	100	100	100	100	100	100	100	93.3	100	100	99.33
	C	100	100	100	100	100	100	100	100	100	100	100