Table 3.
Precision values (%) at 95% recall for the different classifiers.
| SOM | NN | SVC | RF | ADA | ETC | ||
|---|---|---|---|---|---|---|---|
| Euclid | EL-COSMOS | 13.9 | 17.5 | 17.3 | 16.4 | 12.9 | 16.7 |
| Flagship2 | 12.7 | 16.0 | 18.0 | 15.5 | 10.4 | 16.9 | |
| Flagship2 morphology | 9.6 | 17.6 | 16.8 | 15.3 | 11.0 | 14.7 | |
| Euclid + | EL-COSMOS ground | 20.7 | 34.3 | 34.3 | 31.5 | 29.1 | 28.0 |
| Flagship2 ground | 26.1 | 47.9 | 43.5 | 39.3 | 39.7 | 35.6 | |
Notes. The two top rows give the results for training using Euclid photometry only, while morphological data and ground-based photometry, respectively, are used in the bottom rows. The relative uncertainty on all values is ∼6%, estimated from multiple realisations of the training and test sets.
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.