Table 17 Best models from two datasets compared in the end.
Model combination | Dataset | Testing accuracy (%) | AUC (%) |
|---|---|---|---|
ADaDR-22 + Vision transformer | Original (24 Classes) | 89.1 | 97 |
ADaDR-22 + Vision transformer | Refined (14 Classes) | 93.4 | 98.8 |
ADaDR-22 + XGBoost | Original (24 Classes) | 88.3 | 96.5 |
ADaDR-22 + XGBoost | Refined (14 Classes) | 92.2 | 97.8 |
