Table 5 Performance comparison at different parameters \(\alpha\)and \(\tau\).
From: SCBM-Net: a multimodal feature fusion-based dual-channel method for bearing fault diagnosis
α | τ | RE | OI | Accuracy (%) | Precision | Recall | F1-score |
|---|---|---|---|---|---|---|---|
1000 | \({10^{{\text{-6}}}}\) | 0.185133 | 0.339019 | 94.00% | 0.9442 | 0.9400 | 0.9347 |
1000 | \({10^{{\text{-7}}}}\) | 0.184928 | 0.339040 | 94.33% | 0.9508 | 0.9433 | 0.9409 |
1000 | \({10^{{\text{-8}}}}\) | 0.184867 | 0.339051 | 95.33% | 0.9544 | 0.9533 | 0.9516 |
2000 | \({10^{{\text{-6}}}}\) | 0.261097 | 0.282827 | 96.83% | 0.9695 | 0.9683 | 0.9677 |
2000 | \({10^{{\text{-7}}}}\) | 0.260675 | 0.282948 | 97.50% | 0.9726 | 0.9750 | 0.9732 |
2000 | \({10^{{\text{-8}}}}\) | 0.260630 | 0.282974 | 96.50% | 0.9651 | 0.9650 | 0.9643 |
5000 | \({10^{{\text{-6}}}}\) | 0.393508 | 0.090837 | 97.00% | 0.9698 | 0.9700 | 0.9697 |
5000 | \({10^{{\text{-7}}}}\) | 0.393401 | 0.090806 | 96.50% | 0.9660 | 0.9650 | 0.9638 |
5000 | \({10^{{\text{-8}}}}\) | 0.393400 | 0.090808 | 96.67% | 0.9672 | 0.9667 | 0.9662 |
