Learning Beyond Experience: Generalizing to Unseen State Space with Reservoir Computing
Fuente:
arXiv
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866911238430130176 |
|---|---|
| author | Norton, Declan A. Zhang, Yuanzhao Girvan, Michelle |
| author_facet | Norton, Declan A. Zhang, Yuanzhao Girvan, Michelle |
| contents | Machine learning techniques offer an effective approach to modeling dynamical systems solely from observed data. However, without explicit structural priors -- built-in assumptions about the underlying dynamics -- these techniques typically struggle to generalize to aspects of the dynamics that are poorly represented in the training data. Here, we demonstrate that reservoir computing -- a simple, efficient, and versatile machine learning framework often used for data-driven modeling of dynamical systems -- can generalize to unexplored regions of state space without explicit structural priors. First, we describe a multiple-trajectory training scheme for reservoir computers that supports training across a collection of disjoint time series, enabling effective use of available training data. Then, applying this training scheme to multistable dynamical systems, we show that RCs trained on trajectories from a single basin of attraction can achieve out-of-domain generalization by capturing system behavior in entirely unobserved basins. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_05292 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Learning Beyond Experience: Generalizing to Unseen State Space with Reservoir Computing Norton, Declan A. Zhang, Yuanzhao Girvan, Michelle Machine Learning Dynamical Systems Chaotic Dynamics Computational Physics Machine learning techniques offer an effective approach to modeling dynamical systems solely from observed data. However, without explicit structural priors -- built-in assumptions about the underlying dynamics -- these techniques typically struggle to generalize to aspects of the dynamics that are poorly represented in the training data. Here, we demonstrate that reservoir computing -- a simple, efficient, and versatile machine learning framework often used for data-driven modeling of dynamical systems -- can generalize to unexplored regions of state space without explicit structural priors. First, we describe a multiple-trajectory training scheme for reservoir computers that supports training across a collection of disjoint time series, enabling effective use of available training data. Then, applying this training scheme to multistable dynamical systems, we show that RCs trained on trajectories from a single basin of attraction can achieve out-of-domain generalization by capturing system behavior in entirely unobserved basins. |
| title | Learning Beyond Experience: Generalizing to Unseen State Space with Reservoir Computing |
| topic | Machine Learning Dynamical Systems Chaotic Dynamics Computational Physics |
| url | https://arxiv.org/abs/2506.05292 |