Nuclear binding energy predictions using neural networks: Application of the multilayer perceptron

TL;DR

This paper demonstrates that neural networks, specifically multilayer perceptrons, can accurately predict atomic nuclei binding energies, offering a fast alternative to traditional methods by leveraging different architectures and input features.

Contribution

The study introduces the application of multilayer perceptrons with various architectures to predict nuclear binding energies, highlighting the importance of input features for accuracy.

Findings

MLP can predict binding energies reliably.

Different architectures impact prediction accuracy.

Including physical information improves results.

Abstract

In recent years, artificial neural networks and their applications for large data sets have became a crucial part of scientific research. In this work, we implement the Multilayer Perceptron (MLP), which is a class of feedforward artificial neural network (ANN), to predict ground-state binding energies of atomic nuclei. Two different MLP architectures with three and four hidden layers are used to study their effects on the predictions. To train the MLP architectures, two different inputs are used along with the latest atomic mass table and changes in binding energy predictions are also analyzed in terms of the changes in the input channel. It is seen that using appropriate MLP architectures and putting more physical information in the input channels, MLP can make fast and reliable predictions for binding energies of atomic nuclei, which is also comparable to the microscopic energy density functionals.