Speaker
Description
Bayesian methods provide a unifying and powerful framework to quantify uncertainties. Their recent application to R-matrix calculations of nuclear reactions has lent a new level of rigor to R-matrix analyses, accessing new information about R-matrix parameters and assessing the impact of experimental systematics on both estimates of those parameters and R-matrix predictions.
I will begin by reviewing Bayesian methods within the specific R-matrix context. I will argue, using examples, that sampling of the full Bayesian posterior for the R-matrix parameter vector $\vec{\theta}$ provides insights into the formulation of the R-matrix model that are difficult to obtain within a frequentist approach. I will then explain how, with that posterior in hand, it is straightforward to obtain predictions for as-yet-unmeasured reaction data.
Bayesian methods also permit straightforward modeling of experimental imperfections. I will show recent examples of Bayesian R-matrix analyses that treated experimental systematics such as acceptance effects at specific (lab.) angles [1], beam-energy shifts [2] and that examined the impact of inflating the point-to-point uncertainties of different data sets [3].
If time permits I will close by touching on the treatment of theory imperfections within a Bayesian approach.
This work was supported by the U.S. Department of Energy, National Nuclear Security Agency under Award DE- NA0003883, and Office of Science, Office of Nuclear Physics, under Award DE-FG02-93 ER-40756, as well as by the National Science Foundation under grants OAC-2004601 (CSSI program, BAND collaboration).
[1] S. N. Paneru et al., Phys. Rev. C (accepted), arXiv:2211.14641.
[2] D. Odell, C. R. Brune, D. R. Phillips, R. J. deBoer, and S. N. Paneru, Front. Phys. 10:888476 (2022), doi: 10.3389/fphy.2022.888476.
[3] D. Odell, C. R. Brune, D. R. Phillips, Phys. Rev. C 105, 014625 (2022).
