Improving Neutrino Energy Reconstruction with Machine Learning

TL;DR

This paper demonstrates that dense neural networks significantly improve neutrino energy reconstruction accuracy by estimating inaccessible variables, leading to better experimental sensitivity in neutrino physics.

Contribution

The study introduces a neural network-based method that surpasses traditional algorithms in neutrino energy reconstruction, reducing uncertainties and enhancing experimental performance.

Findings

Energy resolution improved by up to a factor of two.

Equivalent to a 10-30% increase in experimental exposure.

Neural networks effectively estimate inaccessible kinematic variables.

Abstract

Faithful energy reconstruction is foundational for precision neutrino experiments like DUNE, but is hindered by uncertainties in our understanding of neutrino--nucleus interactions. Here, we demonstrate that dense neural networks are very effective in overcoming these uncertainties by estimating inaccessible kinematic variables based on the observable part of the final state. We find improvements in the energy resolution by up to a factor of two compared to conventional reconstruction algorithms, which translates into an improved physics performance equivalent to a 10-30% increase in the exposure.