Translation Invariance-Based Deep Learning for Rotating Machinery Diagnosis

Table 2

Performance comparison of different methods with the motor bearing dataset (unit: %).

Methods	Ratio of training samples (%)	Load (hp)	Number of health conditions	Testing accuracy (%)
SVM [23]	75	0	10	88.90
TR-LDA [24]	10	3	10	92.50
EDAEs [5]	66.7	0	12	97.18
2S_LM [1]	10		10	99.66
Developed	10		10	99.71