Open Access
Table 5.
Results from Planck data on two different sky masks, using Gaussian NNs, SRoll2-retrained NN models, and the empirical Cℓ-based likelihood presented in Pagano et al. (2020).
| Predictions on Planck SRoll2 data | ||||||
|---|---|---|---|---|---|---|
| 143 + 100 GHz | 143 + 100 GHz | 143 × 100 GHz | ||||
| Gaussian training | SRoll2 retraining | Cℓ likelihood | ||||
| fsky | τNN | σ(τNN) | τNN | σ(τNN) | τ | σ(τ) |
| 50% | 0.0588 | 0.0063 | 0.0579 | 0.0082 | 0.0566 | 0.0062 |
| 60% | 0.0593 | 0.0059 | 0.0583 | 0.0078 | 0.0577 | 0.0054 |
Notes. The NN results are averaged over 100 models, and σ(τNN) is computed from 10 000 simulations with input τ = 0.058.
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.