Table 2 Diagnostic accuracy of different methods.
From: Intelligent fault diagnosis based on multi-source information fusion and attention-enhanced networks
| Â | Method | F1 score | Accuracy | Parameters | Inference time (ms) |
|---|---|---|---|---|---|
Single-source information | CNN-LSTM | 71.25% | 71.55% | 133k | 4.26 |
IFS-FACNN | 89.32% | 89.41% | 79k | 140.4 | |
SACL | 97.05% | 97.05% | 133k | 5.83 | |
Multi-source information | MH1DCNNs | 98.14% | 98.14% | 4199k | 10.99 |
M-IPISincNet | 98.15% | 98.25% | 66k | 59.99 | |
MSICNNs | 99.08% | 99.09% | 5039k | 1297.55 | |
MIRCA | 99.61% | 99.61% | 21621k | 27.62 |
