Table C.1.
Metrics for the EWS, with the paired labels approach.
| CSMR | CCR | DLNN | nnpz | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| NMAD | fout | bias | NMAD | fout | bias | NMAD | fout | bias | NMAD | fout | bias | ||
| C16 | z | 0.07 | 24% | −0.03 | 0.06 | 20% | −0.03 | 0.07 | 22% | −0.03 | 0.06 | 18% | −0.04 |
| M⋆ | 0.25 | 20% | −0.10 | 0.24 | 20% | −0.08 | 0.25 | 21% | −0.09 | 0.23 | 17% | −0.01 | |
| SFR | 0.82 | 34% | −0.05 | 0.80 | 33% | −0.04 | 0.83 | 35% | −0.11 | 0.70 | 28% | 0.16 | |
| C25 | z | 0.07 | 24% | −0.03 | 0.06 | 20% | −0.03 | 0.07 | 25% | −0.03 | 0.06 | 18% | −0.04 |
| M⋆ | 0.25 | 21% | −0.11 | 0.24 | 21% | −0.08 | 0.25 | 21% | −0.09 | 0.22 | 17% | −0.05 | |
| SFR | 0.82 | 34% | −0.08 | 0.80 | 33% | −0.07 | 0.82 | 34% | 0.03 | 0.69 | 28% | 0.14 | |
| EDF | z | 0.06 | 21% | −0.01 | 0.05 | 17% | −0.01 | 0.07 | 22% | −0.03 | 0.05 | 16% | −0.02 |
| M⋆ | 0.21 | 22% | −0.03 | 0.19 | 20% | −0.05 | 0.22 | 23% | −0.08 | 0.19 | 17% | 0.01 | |
| SFR | 0.79 | 30% | −0.06 | 0.75 | 28% | −0.09 | 0.81 | 31% | −0.12 | 0.65 | 25% | −0.09 | |
Notes. The leftmost column refers to the training (reference) sample, i.e., C25 means a model trained with features and labels from the C25 simulated auxiliary field. All the models are then tested on features from the EWS simulation and ground truth labels. The reported metrics are the ones presented in Sect. 3.6. We are not reporting the paired labels EWS row as it is the same as the mixed labels T.lab. Wide-Wide case in Figs. 5–7 and Table 5. M⋆ refers to log10(M⋆/M⊙), SFR to log10 (SFR/M⊙ yr−1).
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.