Table 8 Summary of evaluation metrics used for classification, uncertainty quantification, and cross-domain adaptation
Category | Metrics | Significance |
|---|---|---|
Classification | AUROC, Accuracy, F1-score | Discriminative power, overall correctness, class balance |
Uncertainty quantification | ECE, Brier Score, NLL | Calibration, probabilistic reliability, confidence quality |
Cross-domain adaptation | ΔAcc, ΔAUROC | Robustness gains under distributional shift |
