Translation Invariance-Based Deep Learning for Rotating Machinery Diagnosis

Table 3

Classification comparison of worm gear dataset.

Methods	Percentage of training samples (%)	Load ()	Number of health conditions	Test accuracy (%)
SFDSC [25]	50	0	4	96.67
SFDSC [25]	50	6	4	88.57
Wavelet-SSAE	10	0/6	4	96.94
Wavelet-SSAE	25	0/6	4	98.16
DCWT-SSAE	10	0/6	4	99.92
DCWT-SSAE	25	0/6	4	99.98