Shell quenching in nuclear charge radii based on Monte Carlo dropout Bayesian neural network
Oct 21, 2024Citations per year
0 Citations
Abstract: (arXiv)
Charge radii can be generally used to encode the information about various fine structures of finite nuclei. In this work, a constructed Bayesian neural network based on Monte Carlo dropout approach is proposed to accurately describe the charge radii of nuclei with proton number and mass number . More motivated underlying mechanisms are incorporated into this combined model except for the basic building blocks with the specific proton and neutron numbers, which naturally contains the pairing effect, isospin effect, shell closure effect associated to the Casten factor , valence neutrons, valence protons, quadrupole deformation , high order hexadecapole deformation , and the local ``abnormal" shape staggering effect of Hg. To avoid the over-fitting puzzle along , , and isotopic chains, the modified Casten factor is tentatively introduced into the input structure parameter sets. We have successfully demonstrated the ability of the Monte Carlo dropout Bayesian neural network (MC-dropout BNN) model to significantly increase the accuracy in predicting the nuclear charge radii. The standard root-mean-square deviation falls into fm for the training data and fm for the validation data with the modified Casten factor input. Meanwhile, the shell closure effect of nuclear charge radii can be reproduced remarkably well.Note:
- 13 pages, 3 figures and 2 tables, Submitted to Physics Letters B
References(140)
Figures(3)
- [1]
- [2]
- [3]
- [4]
- [5]
- [6]
- [7]
- [8]
- [9]
- [10]
- [11]
- [12]
- [13]
- [14]
- [15]
- [15]
- [16]
- [17]
- [18]
- [19]
- [20]
- [21]
- [22]
- [23]
- [24]