TY  - RPRT
TI  - A multi-scale loss formulation for learning a probabilistic model with proper score optimisation
AU  - Simon Lang
AU  - Martin Leutbecher
AU  - Pedro Maciel
PY  - 2025
UR  - https://arxiv.org/abs/2506.10868
ID  - 2506.10868
ER  -