Novel Bayesian neural network based approach for nuclear charge radii

Dong, Xiao-Xu; An, Rong; Lu, Jun-Xu; Geng, Li-Sheng

doi:10.1103/physrevc.105.014308

Cited by 44 publications

(23 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Therefore, it can be seen that although KRR method is a powerful machine learning method, a microscopic model which can provide a better description of the nuclear charge radius is still needed. Note that the Bayesian neural network has also been applied to study the nuclear charge radii for the Ca and K isotopes recently [53].…”

Section: Resultsmentioning

confidence: 99%

“…In recent years, machine learning (ML) has been employed to further improve the accuracies of nuclear models due to its powerful and convenient inference abilities. Various ML approaches have been adopted to improve the description of the nuclear charge radii, e.g., the feedforward neural network [48,49], the Bayesian neural network approach [50][51][52][53], etc. By training the ML network with the deviations between experimental and calculated charge radii, ML approaches can reduce the corresponding rms deviations significantly to about 0.02 fm.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Improved phenomenological nuclear charge radius formulae with kernel ridge regression

Ma,

Zhang

2022

Preprint

View full text Add to dashboard Cite

The kernel ridge regression (KRR) method with Gaussian kernel is used to improve the description of the nuclear charge radius by several phenomenological formulae. The widely used A 1/3 , N 1/3 and Z 1/3 formulae, and their improved versions by considering the isospin dependence are adopted as examples. The parameters in these six formulae are refitted using the Levenberg-Marquardt method, which give better results than the previous ones. The radius for each nucleus is predicted with the KRR network, which is trained with the deviations between experimental and calculated nuclear charge radii. For each formula, the resultant root-mean-square deviations of 884 nuclei with proton number Z ≥ 8 and neutron number N ≥ 8 can be reduced to about 0.017 fm after considering the modification of the KRR method. The extrapolation ability of the KRR method for the neutron-rich region is examined carefully and compared with the radial basis function method. It is found that the improved nuclear charge radius formulae by KRR method can avoid the risk of overfitting and have a good extrapolation ability. The influence of the ridge penalty term on the extrapolation ability of the KRR method is also discussed. At last, the nuclear charge radii of several recently observed K and Ca isotopes have been analyzed.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Improved phenomenological nuclear charge radius formulae with kernel ridge regression

Ma,

Zhang

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Nevertheless, it is well known that for small data sets, engineered features (in addition to these "fundamental features") , which encode important information (priors) about the system under investigation, can play an invaluable role in enhancing the capacity of ANNs. Such a technique is widely used in nuclear physics studies (see, e.g., [38,39,54,55]). In Ref.…”

Section: A Artificial Neural Networkmentioning

confidence: 99%

“…In recent years, artificial neural networks (ANNs), as one of the most powerful machine learning methods, have been successfully applied in nuclear physics studies [9,32], e.g., binding energies [33][34][35][36][37] , charge radii [38][39][40][41], α-decay half-lives [42] , β-decay half-lives [43], and fission fragment yields [44][45][46].…”

Section: Introductionmentioning

confidence: 99%

Nuclear binding energies in artificial neural networks

Zeng¹,

Yin²,

Dong³

et al. 2022

Preprint

View full text Add to dashboard Cite

The binding energy (BE) or mass is one of the most fundamental properties of an atomic nucleus. Precise binding energies are vital inputs for many nuclear physics and nuclear astrophysics studies. However, due to the complexity of atomic nuclei and of the non-perturbative strong interaction, up to now, no conventional physical model can describe nuclear binding energies with a precision below 0.1 MeV, the accuracy needed by nuclear astrophysical studies. In this work, artificial neural networks (ANNs), the so called "universal approximators", are used to calculate nuclear binding energies. We show that the ANN can describe all the nuclei in AME2020 with a root-mean-square deviation (RMSD) around 0.2 MeV, which is better than the best macroscopic-microscopic models, such as FRDM and WS4. The success of the ANN is mainly due to the proper and essential input features we identify, which contain the most relevant physical information, i.e., shell, paring, and isospin-asymmetry effects. We show that the well-trained ANN has excellent extrapolation ability and can predict binding energies for those nuclei so far inaccessible experimentally. In particular, we highlight the important role played by "feature engineering" for physical systems where data are relatively scarce, such as nuclear binding energies.

show abstract

“…The nuclear charge radius along the isotope chains became a hot topic in recent discussion. It's being one of the most fundamental properties of a nucleus, plays vital role in our understanding of complex dynamics of atomic nuclei [1]. The knowledge of nuclear size plays an important role not only in understanding new physics beyond standard model (SM) but also serving as input quantities in astrophysical study [2].…”

Section: Introductionmentioning

confidence: 99%

Role of Isovector-Isoscalar Coupling on Charge Radius of Heavy and Superheavy Nuclei

Liliani

2022

SPEKTRA

View full text Add to dashboard Cite

We have investigated the effect of the isovector-isoscalar coupling on the finite nuclei and nuclear matter properties, the neutron skin thickness of 208Pb, and the charge radius on heavy and superheavy nuclei calculated by the relativistic mean-field (RMF) model. In this work, we generates two parameter sets, i.e., PTE16 and PTE31. The numbers 16 and 31 denote the isovector-isoscalar coupling terms, while T and E denote the tensor coupling and electromagnetic exchange terms, respectively. We found that PTE16 and PTE31 are compatible with the constraints obtained by R. Essick, et al., arXiv: 2102.10074v1 [nucl-th] (2021). We also found that the increase of the isovector-isoscalar coupling terms gives a significant effect on the binding energy and the charge radius on heavy nuclei except for the charge radius of 208Pb. Increased of the isovector-isoscalar coupling terms make the values of charge radius prediction increase too, but vice versa for the neutron skin thickness and nuclear matter prediction. PTE31 yields symmetry energy J = 31.521 MeV, slope L = 57.643 MeV, and neutron skin thickness = 0.21419 fm. While the β2 correction (for deform nuclei) does not always give a significant effect on the charge radius.

show abstract

Novel Bayesian neural network based approach for nuclear charge radii

Cited by 44 publications

References 50 publications

Improved phenomenological nuclear charge radius formulae with kernel ridge regression

Improved phenomenological nuclear charge radius formulae with kernel ridge regression

Nuclear binding energies in artificial neural networks

Role of Isovector-Isoscalar Coupling on Charge Radius of Heavy and Superheavy Nuclei

Contact Info

Product

Resources

About