Combining machine learning and data assimilation to forecast dynamical systems from noisy partial observations
Author:
Affiliation:
1. School of Mathematics and Statistics, The University of Sydney, Sydney, NSW 2006, Australia
2. Institute of Mathematics, University of Potsdam, D-14476 Potsdam, Germany
Funder
Deutsche Forschungsgemeinschaft
Publisher
AIP Publishing
Subject
Applied Mathematics,General Physics and Astronomy,Mathematical Physics,Statistical and Nonlinear Physics
Link
https://aip.scitation.org/doi/pdf/10.1063/5.0066080
Reference67 articles.
1. Solving high-dimensional partial differential equations using deep learning
2. Uniformly accurate machine learning-based hydrodynamic models for kinetic equations
3. Hidden physics models: Machine learning of nonlinear partial differential equations
4. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
5. Earth System Modeling 2.0: A Blueprint for Models That Learn From Observations and Targeted High‐Resolution Simulations
Cited by 19 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献
1. CGNSDE: Conditional Gaussian neural stochastic differential equation for modeling complex systems and data assimilation;Computer Physics Communications;2024-11
2. A Koopman–Takens Theorem: Linear Least Squares Prediction of Nonlinear Time Series;Communications in Mathematical Physics;2024-05
3. Artificial Intelligence and Its Application in Numerical Weather Prediction;Russian Meteorology and Hydrology;2024-04
4. A Machine Learning Augmented Data Assimilation Method for High‐Resolution Observations;Journal of Advances in Modeling Earth Systems;2024-01
5. Autoregressive With Slack Time Series Model for Forecasting a Partially-Observed Dynamical Time Series;IEEE Access;2024
1.学者识别学者识别
2.学术分析学术分析
3.人才评估人才评估
"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370
www.globalauthorid.com
TOP
Copyright © 2019-2024 北京同舟云网络信息技术有限公司 京公网安备11010802033243号 京ICP备18003416号-3