Table 8 Testing results of classification accuracy of CWRU dataset.

From: A hybrid approach combining deep learning and signal processing for bearing fault diagnosis under imbalanced samples and multiple operating conditions

Approaches

Classififcation accuray

SVC

69.52%

RFC

71.20%

EMD-GA-ANFIS(R)

77.15%

XGBoost

84.04%

SOM

86.40%

F-ANFIS

87.70%

MF-SVMs

88.90%

LightGBM

89.09%

DPSON

90.60%

EMD-GA-ANFIS(S)

91.33%

Compact 1DCNN

93.20%

CNN

93.54%

DNN

94.40%

SAE

94.40%

Proposd CWT-MCA-MHAC

97.52%

Back to article page