期刊
GEOPHYSICAL RESEARCH LETTERS
卷 48, 期 8, 页码 -出版社
AMER GEOPHYSICAL UNION
DOI: 10.1029/2020GL092133
关键词
data-driven modeling; machine learning; toy models
Researchers introduced a new toy model of atmospheric dynamics and found that traditional neural networks can lead to unstable trajectories in this model. Training on different learning samples based on Latin Hypercube Sampling can solve this issue and significantly influence the stability of dynamical systems driven by neural networks.
The development of atmospheric parameterizations based on neural networks is often hampered by numerical instability issues. Previous attempts to replicate these issues in a toy model have proven ineffective. We introduce a new toy model for atmospheric dynamics, which consists in an extension of the Lorenz'63 model to a higher dimension. While feedforward neural networks trained on a single orbit can easily reproduce the dynamics of the Lorenz'63 model, they fail to reproduce the dynamics of the new toy model, leading to unstable trajectories. Instabilities become more frequent as the dimension of the new model increases, but are found to occur even in very low dimension. Training the feedforward neural network on a different learning sample, based on Latin Hypercube Sampling, solves the instability issue. Our results suggest that the design of the learning sample can significantly influence the stability of dynamical systems driven by neural networks.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据