Eviatar Escolar Bach
@eviatarbach.bsky.social
Lecturer (Asst. Prof.) in meteorology and math. & stats | Data assimilation, climate, ML, applied dynamical systems, predictability, inverse problems, climate politics
She/her 🏳️⚧️
https://eviatarbach.com
She/her 🏳️⚧️
https://eviatarbach.com
In synthetic experiments with prescribed signal-to-noise ratios, we find that new scoring rule–based estimators agree with the classical ones on whether the signal-to-noise ratio is anomalous or not. In application to the North Atlantic Oscillation, the results are more equivocal. (3/3)
November 3, 2025 at 10:17 AM
In synthetic experiments with prescribed signal-to-noise ratios, we find that new scoring rule–based estimators agree with the classical ones on whether the signal-to-noise ratio is anomalous or not. In application to the North Atlantic Oscillation, the results are more equivocal. (3/3)
We show that the Nesterov acceleration indeed improves the performance of EKI on a variety of inverse problems, at no extra computational cost!
(4/4)
(4/4)
January 16, 2025 at 5:00 PM
We show that the Nesterov acceleration indeed improves the performance of EKI on a variety of inverse problems, at no extra computational cost!
(4/4)
(4/4)