Table 3.
Relative 1σ errors (in percent) from NDE for a single target diagram.
| 103 deg2 | 104 deg2 | 105 deg2 | Infinite deg2 | |||||
|---|---|---|---|---|---|---|---|---|
| Ωm | σ8 | Ωm | σ8 | Ωm | σ8 | Ωm | σ8 | |
| D1 | 15.2 | 10.0 | 9.6 | 5.6 | 4.1 | 2.1 | 0.8 | 0.5 |
| D2 | 17.6 | 13.4 | 11.2 | 7.2 | 7.9 | 5.0 | 4.1 | 2.1 |
| D3 | 12.2 | 5.5 | 5.3 | 2.1 | 2.0 | 0.8 | 0.4 | 0.2 |
| D4 | 15.0 | 10.5 | 7.6 | 4.7 | 4.0 | 2.3 | 2.4 | 0.9 |
Notes. Separate NDE and regressor are trained for each dataset (D1–D4) and each survey area. The results are computed for the same XOD test target, always of the appropriate survey size, simulated with the γ evolution parameters fixed at the fiducial values and with the dn/dz selection. This choice allowed us to use it fairly for the ML model trained on each of our four datasets.
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.