Enregistré dans:
| Auteurs principaux: | , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | https://arxiv.org/abs/2509.14123 |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866916954945617920 |
|---|---|
| author | Liverani, Lorenzo Steynberg, Matthys Zuazua, Enrique |
| author_facet | Liverani, Lorenzo Steynberg, Matthys Zuazua, Enrique |
| contents | We present Hybrid-Cooperative Learning (HYCO), a hybrid modeling framework that iteratively integrates physics-based and data-driven models through a mutual regularization mechanism. Unlike traditional approaches that impose physical constraints directly on synthetic models, HYCO treats the physical and synthetic components as co-trained agents: the physical and synthetic models are nudged toward agreement, while the synthetic model is enhanced to better fit the available data. This cooperative learning scheme is naturally parallelizable and improves robustness to noise as well as to sparse or heterogeneous data. Extensive numerical experiments on both static and time-dependent problems demonstrate that HYCO outperforms classical physics-based and data-driven methods, recovering accurate solutions and model parameters even under ill-posed conditions. The method also admits a natural game-theoretic interpretation, enabling alternating optimization and paving the way for future theoretical developments. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_14123 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | HYCO: Hybrid-Cooperative Learning for Data-Driven PDE Modeling Liverani, Lorenzo Steynberg, Matthys Zuazua, Enrique Optimization and Control Numerical Analysis Analysis of PDEs 00A71, 35R30, 41A27, 68T07, 91A12 We present Hybrid-Cooperative Learning (HYCO), a hybrid modeling framework that iteratively integrates physics-based and data-driven models through a mutual regularization mechanism. Unlike traditional approaches that impose physical constraints directly on synthetic models, HYCO treats the physical and synthetic components as co-trained agents: the physical and synthetic models are nudged toward agreement, while the synthetic model is enhanced to better fit the available data. This cooperative learning scheme is naturally parallelizable and improves robustness to noise as well as to sparse or heterogeneous data. Extensive numerical experiments on both static and time-dependent problems demonstrate that HYCO outperforms classical physics-based and data-driven methods, recovering accurate solutions and model parameters even under ill-posed conditions. The method also admits a natural game-theoretic interpretation, enabling alternating optimization and paving the way for future theoretical developments. |
| title | HYCO: Hybrid-Cooperative Learning for Data-Driven PDE Modeling |
| topic | Optimization and Control Numerical Analysis Analysis of PDEs 00A71, 35R30, 41A27, 68T07, 91A12 |
| url | https://arxiv.org/abs/2509.14123 |