The Hybrid Method of VMD-PSR-SVD and Improved Binary PSO-KNN for Fault Diagnosis of Bearing

Table 2

The comparison of the diagnosis accuracy of bearing between VPS-IBPSOKNN and VS-KNN

The number of training samples	The number of testing samples	Diagnosis method	The number of testing samples with correct diagnosis	Diagnosis accuracy (%)
600	300	VPS-IBPSOKNN	295	98.33
		VS-KNN (K = 1)	268	89.33
		VS-KNN (K = 2)	269	89.67
		VS-KNN (K = 3)	278	92.67
		VS-KNN (K = 4)	277	92.33
		VS-KNN (K = 5)	280	93.33
		VS-KNN (K = 6)	278	92.67
		VS-KNN (K = 7)	279	93.00
		VS-KNN (K = 8)	278	92.67
		VS-KNN (K = 9)	278	92.67
		VS-KNN (K = 10)	278	92.67
		VS-KNN (K = 11)	277	92.33
		VS-KNN (K = 12)	278	92.67
		VS-KNN (K = 13)	277	92.33
		VS-KNN (K = 14)	276	92.00
		VS-KNN (K = 15)	273	91.00