Nuclear masses learned from a probabilistic neural network

TL;DR

This paper employs a probabilistic neural network, specifically a Mixture Density Network, to predict nuclear mass excesses, providing both accurate predictions and meaningful uncertainty quantification, and demonstrates improved extrapolation with physical information.

Contribution

It introduces the use of a Mixture Density Network for nuclear mass prediction, enhancing accuracy and physical relevance in extrapolations beyond experimental data.

Findings

MDN provides full posterior distributions for mass predictions.

Adding physical information improves accuracy and extrapolation.

Model successfully extrapolates beyond measured data.

Abstract

Machine learning methods and uncertainty quantification have been gaining interest throughout the last several years in low-energy nuclear physics. In particular, Gaussian processes and Bayesian Neural Networks have increasingly been applied to improve mass model predictions while providing well-quantified uncertainties. In this work, we use the probabilistic Mixture Density Network (MDN) to directly predict the mass excess of the 2016 Atomic Mass Evaluation within the range of measured data, and we extrapolate the inferred models beyond available experimental data. The MDN not only provides mean values but also full posterior distributions both within the training set and extrapolated testing set. We show that the addition of physical information to the feature space increases the accuracy of the match to the training data as well as provides for more physically meaningful extrapolations beyond the the limits of experimental data.