Table 6 Comparison of classification accuracy of PADERBORN dataset.

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

Approaches

Classification

CNN

68.75%

TCN-LSTM

78.56%

CNN-attention

95.09%

CNN-LSTM

96.43%

LSTM-attention

96.88%

MAC-MHA-Net

97.52%

Back to article page