Physics Informed Bayesian Machine Learning of Sparse and Imperfect Nuclear Data
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866915768551079936 |
|---|---|
| author | Liu, Jiaming Su, Yang Shu, N. C. Chen, Y. J. Pei, J. C. |
| author_facet | Liu, Jiaming Su, Yang Shu, N. C. Chen, Y. J. Pei, J. C. |
| contents | The prevailing data-driven machine learning has been plagued by the absence of physics knowledge and the scarcity of data. We implement the physics-model informed prior into Bayesian machine learning to evaluate the energy dependence of independent fission product yields, which are crucial for advanced nuclear energy applications but only sparse and imperfect experimental data are available. The informative prior is the posterior after learning the generated data from fission models. Furthermore, cumulative fission yields are included as a physical constraint via a conversion matrix to provide augmented energy dependence. Our work demonstrated a truly Bayesian machine learning by incorporating comprehensive physics knowledges as a powerful tool to exploit the sparse but expensive nuclear data. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_01808 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Physics Informed Bayesian Machine Learning of Sparse and Imperfect Nuclear Data Liu, Jiaming Su, Yang Shu, N. C. Chen, Y. J. Pei, J. C. Nuclear Theory The prevailing data-driven machine learning has been plagued by the absence of physics knowledge and the scarcity of data. We implement the physics-model informed prior into Bayesian machine learning to evaluate the energy dependence of independent fission product yields, which are crucial for advanced nuclear energy applications but only sparse and imperfect experimental data are available. The informative prior is the posterior after learning the generated data from fission models. Furthermore, cumulative fission yields are included as a physical constraint via a conversion matrix to provide augmented energy dependence. Our work demonstrated a truly Bayesian machine learning by incorporating comprehensive physics knowledges as a powerful tool to exploit the sparse but expensive nuclear data. |
| title | Physics Informed Bayesian Machine Learning of Sparse and Imperfect Nuclear Data |
| topic | Nuclear Theory |
| url | https://arxiv.org/abs/2602.01808 |