Table 6.
Precision scores of the models considered in this work (as a percentage) for the domain adaptation task considered in scenario 1 (best values are highlighted in bold).
| Domain Shift | RNN-WDCNN | WDCNN | SRDCNN | FFT-SVM | FFT-MLP |
|---|---|---|---|---|---|
| A to B | 100.00 (± 0.00) | 99.89 (± 0.07) | 97.89 (± 0.34) | 94.97 (± 0.24) | 90.07 (± 0.60) |
| A to C | 99.25 (± 0.72) | 97.86 (± 1.01) | 88.66 (± 4.66) | 81.90 (± 1.02) | 79.29 (± 0.63) |
| B to A | 95.92 (± 0.48) | 95.34 (± 0.33) | 96.32 (± 0.65) | 95.63 (± 0.62) | 94.69 (± 0.24) |
| B to C | 99.92 (± 0.09) | 99.85 (± 0.09) | 97.71 (± 1.95) | 87.59 (± 0.73) | 89.68 (± 0.44) |
| C to A | 83.46 (± 1.29) | 80.41 (± 1.77) | 77.03 (± 6.51) | 75.86 (± 1.24) | 78.95 (± 0.64) |
| C to B | 99.17 (± 1.09) | 96.24 (± 1.06) | 93.17 (± 6.03) | 85.27 (± 0.61) | 83.12 (± 0.49) |
| Average | 96.29 (± 0.61) | 94.93 (± 0.72) | 91.80 (± 3.36) | 86.87 (± 0.74) | 85.97 (± 0.51) |