PAC Learnability of nuclear masses

03/29/2019
by   Andrea Idini, et al.
0

After more than 80 years from the seminal work of Weizsäcker and the liquid drop model of the atomic nucleus, theoretical errors over nuclear masses (∼ MeV) are order of magnitudes larger than experimental ones (≲ keV). Predicting the mass of atomic nuclei is with precision is extremely challenging due to the non--trivial many--body interplay of protons and neutrons in nuclei, and the complex nature of the nuclear strong force. This paper argues that the arduous development of nuclear physics in the passed century is due to the exploration of a system on the limit of the knowledgeable, defined within the statistical theory of learning.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
05/08/2023

Predicting nuclear masses with product-unit networks

Accurate estimation of nuclear masses and their prediction beyond the ex...
research
11/11/2022

An introduction to computational complexity and statistical learning theory applied to nuclear models

The fact that we can build models from data, and therefore refine our mo...
research
06/09/2023

NuCLR: Nuclear Co-Learned Representations

We introduce Nuclear Co-Learned Representations (NuCLR), a deep learning...
research
10/06/2018

Deep Learning: Extrapolation Tool for Ab Initio Nuclear Theory

Ab initio approaches in nuclear theory, such as the No-Core Shell Model ...
research
03/21/2018

Activation cross-section data for alpha-particle induced nuclear reactions on natural ytterbium for some longer lived radioisotopes

Additional experimental cross sections were deduced for the long half-li...
research
02/24/2020

Trees and Forests in Nuclear Physics

We present a detailed introduction to the decision tree algorithm using ...

Please sign up or login with your details

Forgot password? Click here to reset