New approach for the quantification of uncertainties in reaction modeling via data-driven multi-objective optimization
Nuclear Theory
2025-07-10 v1
Abstract
We introduce a new multi-objective optimization approach to determine uncertainty-quantified nuclear reaction parameters in the Hauser-Feshbach framework. By simultaneously accounting for all available data across multiple reaction channels we capture parameter correlations and estimate data-driven uncertainties. We implement in the Ni-Ge region yielding uncertainty-quantified model parameters for both stable and unstable isotopes. We estimate resonance spacings for nuclei beyond experimental reach and validate our method by calculating a known cross-section outside our optimization region.
Cite
@article{arxiv.2507.06370,
title = {New approach for the quantification of uncertainties in reaction modeling via data-driven multi-objective optimization},
author = {N. Dimitrakopoulos and G. Perdikakis and F. Montes and P. Gastis and S. A. Kuvin and H. Y. Lee and P. Tsintari and A. V. Voinov},
journal= {arXiv preprint arXiv:2507.06370},
year = {2025}
}