Salvato in:
Dettagli Bibliografici
Autori principali: Adrian, Melissa, Sanz-Alonso, Daniel, Willett, Rebecca
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:https://arxiv.org/abs/2603.20891
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866910062642987008
author Adrian, Melissa
Sanz-Alonso, Daniel
Willett, Rebecca
author_facet Adrian, Melissa
Sanz-Alonso, Daniel
Willett, Rebecca
contents Data assimilation algorithms estimate the state of a dynamical system from partial observations, where the successful performance of these algorithms hinges on costly parameter tuning and on employing an accurate model for the dynamics. This paper introduces a framework for jointly learning the state, dynamics, and parameters of filtering algorithms in data assimilation through a process we refer to as auto-differentiable filtering. The framework leverages a theoretically motivated loss function that enables learning from partial, noisy observations via gradient-based optimization using auto-differentiation. We further demonstrate how several well-known data assimilation methods can be learned or tuned within this framework. To underscore the versatility of auto-differentiable filtering, we perform experiments on dynamical systems spanning multiple scientific domains, such as the Clohessy-Wiltshire equations from aerospace engineering, the Lorenz-96 system from atmospheric science, and the generalized Lotka-Volterra equations from systems biology. Finally, we provide guidelines for practitioners to customize our framework according to their observation model, accuracy requirements, and computational budget.
format Preprint
id arxiv_https___arxiv_org_abs_2603_20891
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Auto-differentiable data assimilation: Co-learning of states, dynamics, and filtering algorithms
Adrian, Melissa
Sanz-Alonso, Daniel
Willett, Rebecca
Machine Learning
Signal Processing
Dynamical Systems
Data assimilation algorithms estimate the state of a dynamical system from partial observations, where the successful performance of these algorithms hinges on costly parameter tuning and on employing an accurate model for the dynamics. This paper introduces a framework for jointly learning the state, dynamics, and parameters of filtering algorithms in data assimilation through a process we refer to as auto-differentiable filtering. The framework leverages a theoretically motivated loss function that enables learning from partial, noisy observations via gradient-based optimization using auto-differentiation. We further demonstrate how several well-known data assimilation methods can be learned or tuned within this framework. To underscore the versatility of auto-differentiable filtering, we perform experiments on dynamical systems spanning multiple scientific domains, such as the Clohessy-Wiltshire equations from aerospace engineering, the Lorenz-96 system from atmospheric science, and the generalized Lotka-Volterra equations from systems biology. Finally, we provide guidelines for practitioners to customize our framework according to their observation model, accuracy requirements, and computational budget.
title Auto-differentiable data assimilation: Co-learning of states, dynamics, and filtering algorithms
topic Machine Learning
Signal Processing
Dynamical Systems
url https://arxiv.org/abs/2603.20891