Table 1 Cross-validation % accuracy (average ± standard deviation) obtained using each algorithm and source dataset to re-train the CNN
| Â | CiCC | ISRUC | MASS | SHHS | WSC | MrOS |
|---|---|---|---|---|---|---|
Head Re-train | 78.1 ± 6.3 | 75.9 ± 6.8 | 75.5 ± 6.9 | 76.0 ± 6.5 | 76.5 ± 6.4 | 76.2 ± 8.9 |
Subspace alignment | 53.7 ± 14.1 | 48.0 ± 12.6 | 43.3 ± 11.8 | 45.9 ± 11.3 | 45.0 ± 12.0 | 36.7 ± 12.9 |
CORAL | 78.0 ± 6.3 | 76.1 ± 6.3 | 76.8 ± 6.1 | 75.5 ± 6.3 | 76.6 ± 6.0 | 77.1 ± 6.4 |
Per-Class CORAL | 78.0 ± 5.7 | 75.7 ± 6.8 | 76.4 ± 5.9 | 75.6 ± 6.2 | 76.0 ± 6.1 | 77.2 ± 5.9 |
DDC | 75.7 ± 8.3 | 77.7 ± 6.7 | 77.4 ± 8.5 | 78.8 ± 7.5 | 75.9 ± 9.1 | 79.0 ± 6.3 |