Electron-Nucleus Cross Sections from Transfer Learning
Transfer learning allows a deep neural network (DNN) trained on one type of data to be adapted for new problems with limited information. We propose to use the transfer learning technique in physics. The DNN learns the details of one process, and after fine-tuning, it makes predictions for related processes. We consider the DNNs, trained on inclusive electron-carbon scattering data, and show that after fine-tuning, they accurately predict cross sections for electron interactions with nuclear targets ranging from helium-3 to iron.
Authors
- Beata E. Kowal (ORCID: https://orcid.org/0000-0003-3646-1653)
- J. T. Sobczyk (ORCID: https://orcid.org/0000-0003-4991-2919)
- Hemant Prasad (ORCID: https://orcid.org/0009-0003-1897-9616)
- Rwik Dharmapal Banerjee (ORCID: https://orcid.org/0000-0003-3639-7532)
- J. L. Bonilla (ORCID: https://orcid.org/0009-0009-3240-1494)
- Artur M. Ankowski (ORCID: https://orcid.org/0000-0003-4073-8686)
- Karl J. Friston
Institutions
- University of Wrocław (PL)
Publication Details
- Journal
- Physical Review Letters
- Published
- 2025-07-14
- DOI
- https://doi.org/10.1103/zxv6-22tz
- Citations
- 3
- Primary Topic
- Nuclear Physics and Applications
- Type
- article
- Field-Weighted Citation Impact
- 5.65
Funders
- Narodowe Centrum Nauki
- Uniwersytet Wrocławski