KPCA and AE Based Local-Global Feature Extraction Method for Vibration Signals of Rotating Machinery

Table 8

Classification comparison of the bearing dataset.

Methods	Description	No. of classes	Training samples (%)	Testing accuracy (%)
Proposed	SFS-KPCA-AE + KNN	10	25	99.93
Proposed	SFS-KPCA-AE + SVM	10	25	99.97
[36]	EEMD-ICD + SVM	6	25	98.22
[37]	EMD-WKLFDA + SVM	10	40	98.80
[38]	Bispectrum features + SVM	4	50	96.98
[39]	SDA-DNN	10	25	92.35
[40]	GRU-NP-DAE	10	75	99.22
[41]	NSAE-LCN	10	25	99.92