Open Access
Table 4.
Averaged statistics across all individually trained NLL loss and CPRS models.
| Ensembles | Num. Gaussians | CRPS loss | NLL loss | Fraction of outliers | Accuracy |
|---|---|---|---|---|---|
| η | σNMAD | ||||
| Models trained on CRPS loss | 132 | 0.145 | −0.295 | 0.121 | 0.068 |
| Model trained on NLL loss | 156 | 0.146 | −0.383 | 0.12 | 0.065 |
| All combined | 288 | 0.143 | −0.361 | 0.116 | 0.066 |
Notes. The table includes aggregated information for the combined ensemble of all NLL and CRPS models. Furthermore, the fraction of outliers and accuracy are listed for the best-performing ensembles.
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.