High Precision Binding Energies from Physics Informed Machine Learning

Abstract

Twelve physics-informed machine learning models have been trained to model binding energy residuals. Our approach begins with determining the difference between measured experimental binding energies and three different mass models. Then four machine learning approaches are used to train on each energy difference. The most successful ML technique, both in interpolation and extrapolation, is the least squares boosted ensemble of trees. The best model resulting from that technique utilizes eight physical features to model the difference between experimental atomic binding energy values in AME 2012 and the Duflo Zuker mass model. This resulted in a model that fit the training data with a standard deviation of 17 keV and that has a standard deviation of 92 keV when compared all of the values in the AME 2020. The extrapolation capability of each model is discussed, and the accuracy of predicting new mass measurements has also been tested.

Version 2 contains edits including those based on referee feedback

References in corpus (10)

• Surface diffuseness correction in global mass formula
• Nuclear masses learned from a probabilistic neural network
• Nuclear masses in extended kernel ridge regression with odd-even effects
• Nuclear mass predictions using machine learning models
• Detailed illustration of accuracy of presently used nuclear-mass models
• High-precision mass measurements of neutron deficient silver isotopes probe the robustness of the $N$ = 50 shell closure
• Investigating nuclear structure near $N=32$ and $N=34$: Precision mass measurements of neutron-rich Ca, Ti and V isotopes
• Precision mass measurements in the zirconium region pin down the mass surface across the neutron midshell at $N=66$
• High-precision Penning-trap mass measurements of Cd and In isotopes at JYFLTRAP remove the fluctuations in the two-neutron separation energies
• 2023 Update of the Discovery of Nuclides

Cited by in corpus (1)

• Further exploration of binding energy residuals using machine learning and the development of a composite ensemble model