Global Framework for Emulation of Nuclear Calculations

TL;DR

This paper presents a hierarchical framework combining ab initio calculations and Bayesian neural networks to accurately emulate nuclear properties across isotopic chains, with robust uncertainty quantification and sensitivity analysis.

Contribution

It introduces a novel global emulation framework that integrates physics-based models with machine learning for nuclear property predictions.

Findings

Accurately predicts ground-state energies and charge radii for oxygen isotopes.

Provides robust uncertainty quantification in nuclear property predictions.

Enables global sensitivity analysis of nuclear forces.

Abstract

We introduce a hierarchical framework that combines ab initio many-body calculations with a Bayesian neural network, developing emulators capable of accurately predicting nuclear properties across isotopic chains simultaneously and being applicable to different regions of the nuclear chart. We benchmark our developments using the oxygen isotopic chain, achieving accurate results for ground-state energies and nuclear charge radii, while providing robust uncertainty quantification. Our framework enables global sensitivity analysis of nuclear binding energies and charge radii with respect to the low-energy constants that describe the nuclear force.