PaperPanorama
FAQ

Nuclear Theory·nucl-th

Tuesday·October 10, 2023

11 papers—5 primary·6 cross-listed

All11Primary5Cross-lists6
  1. 01

    Normalizing Flows for Bayesian Posteriors: Reproducibility and Deployment

    Yukari Yamauchi · Landon Buskirk · Pablo Giuliani · Kyle Godbey

    We present a computational framework for efficient learning, sampling, and distribution of general Bayesian posterior distributions. The framework leverages a machine learning approach for the construction of normalizing flows for the general probability distributions typically encountered in Bayesian uncertainty quantification studies. This normalizing flow can map a trivial distribution to a more complicated one and can be stored more efficiently than the empirical distribution samples themselves. Once the normalized flow is trained, it further enables parallelized and uncorrelated sampling of the learned distribution. We demonstrate our framework with three test distributions with strong non-linear correlations, multi-modality, and heavy tails, as well as with a realistic posterior distribution obtained from a Bayesian calibration of a nuclear relativistic mean-field model. The performance of the framework, as well as its relatively simple implementation, positions it as one fundamental cornerstone in the development and deployment of continuous calibration pipelines of physical models and as a key component of future reproducible science workflows.

    nucl-th6 citations
  2. 02

    Importance of physical information on the prediction of heavy-ion fusion cross section with machine learning

    Zhilong Li · Zepeng Gao · Ling Liu · Yongjia Wang · Long Zhu · Qingfeng Li

    In this work, the Light Gradient Boosting Machine (LightGBM), which is a modern decision tree based machine-learning algorithm, is used to study the fusion cross section (CS) of heavy-ion reaction. Several basic quantities (e.g., mass number and proton number of projectile and target) and the CS obtained from phenomenological formula are fed into the LightGBM algorithm to predict the CS. It is found that, on the validation set, the mean absolute error (MAE) which measures the average magnitude of the absolute difference between of the predicted CS and experimental CS is 0.129 by only using the basic quantities as the input, this value is smaller than 0.154 obtained from the empirical coupled channel model. MAE can be further reduced to 0.08 by including an physical-informed input feature. The MAE on the test set (it consists of 280 data points from 18 reaction systems that not included in the training set) is about 0.19 and 0.53 by including and excluding the physical-informed feature, respectively. We further verify the LightGBM predictions by comparing the CS of + obtained from the density-constrained time-dependent Hartree-Fock approach. Our study demonstrates the importance of physical information in predicting fusion cross section of heavy-ion reaction with machine learning.

    nucl-thnucl-exPRC(2024)·17 citations
  3. 03

    Impact of choices for center-of-mass correction energy on the surface energy of Skyrme energy density functionals

    Philippe Da Costa🇫🇷 · Karim Bennaceur🇫🇷 · Jacques Meyer🇫🇷 · Wouter Ryssens🇧🇪 · Michael Bender🇫🇷

    In the framework of nuclear energy density functional (EDF) methods, many nuclear phenomena can be related to the deformation of intrinsic states. Their accurate modeling relies on the correct description of the change of nuclear binding energy with deformation. The two most important contributions to the deformation energy have their origin in shell effects and the surface energy coefficient of nuclear matter. In a first step, we build nine series of parametrizations with a systematically varied surface-energy coefficient a_surf for three frequently-used options for the CM correction (none, one-body term only, full one-body and two-body contributions) combined with three values for the isoscalar effective mass m^*_0/m (0.7, 0.8, 0.85) and analyse how well each of these parametrizations can be adjusted to the properties of spherical nuclei and infinite nuclear matter. In a second step, we performed additional fits without the constraint on surface energy, adding one ``best-fit" parametrization to each of the nine series. We then benchmark these parametrizations to the deformation properties of heavy nuclei by means of three-dimensional Hartree-Fock-Bogoliubov calculations that allow for non-axial and/or non-reflection symmetric configurations. We perform a detailed correlation analysis between surface and volume properties of nuclear matter using the nine series of parametrizations. The best fits out of each series are then benchmarked on the fission barriers of Pu240 and Hg180, as well as on the properties of deformed states at normal and superdeformation for actinides and nuclei in the neutron-deficient Hg region. (see paper for full abstract)

    nucl-thPRC(2024)·7 citations
  4. 04

    Precise neural network predictions of energies and radii from the no-core shell model

    Tobias Wolfgruber · Marco Knöll · Robert Roth

    For light nuclei, ab initio many-body methods such as the no-core shell model are the tools of choice for predictive, high-precision nuclear structure calculations. The applicability and the level of precision of these methods, however, is limited by the model-space truncation that has to be employed to make such computations feasible. We present a universal framework based on artificial neural networks to predict the value of observables for an infinite model-space size based on finite-size no-core shell model data. Expanding upon our previous ansatz of training the neural networks to recognize the observable-specific convergence pattern with data from few-body nuclei, we improve the results obtained for ground-state energies and show a way to handle excitation energies within this framework. Furthermore, we extend the framework to the prediction of converged root-mean-square radii, which are more difficult due to the much less constrained convergence behavior. For all observables robust and statistically significant uncertainties are extracted via the sampling over a large number of network realizations and evaluation data samples.

    nucl-thPRC(2024)·26 citations
  5. 05

    Interactions of -Mesons in Nuclear Matter and with Nuclei

    Horst Lenske🇩🇪

    In--medium interactions of --mesons are investigated microscopically by coupling to and particle--hole excitations of the medium, including elativistic mean--field dynamics by self--consistent scalar and vector fields. The resulting self--energies are transmitted to finite nuclei in local density approximation. Real and imaginary parts of longitudinal and transversal self--energies are discussed. The relation of the present approach to meson cloud models is addressed and an ambiguity is pointed out. Applications to recent data on the in--medium width of mesons scattered on a Niobium target serve to determine unknown in-medium coupling constants. The data are well described by self--energies containing S--wave and P--wave resonances. Exploratory investigations, however, show that the spectroscopic composition of self--energies depends crucially on the near--threshold properties of the width which at present is known only within large error bars. The calculations predict the prevalence of transversal self--energies, implying that vector current conservation is still maintained in the nuclear medium by slightly more than 90\%. Schrödinger--equivalent potentials are derived and scattering lengths and effective range parameters are extracted for the longitudinal and transversal channel. Longitudinal and transversal spectral distributions are discussed and the dependencies on momentum and nuclear density are investigated. Schrödinger--type Nb optical potentials are constructed. Low--energy parameters are determined, are used to study the pole structure of the S--matrix at threshold. The effective range expansion of the omega--nucleus K--matrix led to a Nb bound states with binding energy ~keV but of width ~keV.

    nucl-thEPJA(2023)·7 citations

Affiliations

first authorsco-authorsvia INSPIRE