Machine learning based event classification for the energy-differential measurement of the $^\text{nat}$C(n,p) and $^\text{nat}$C(n,d) reactions

TL;DR

This paper demonstrates that neural networks, trained on voxelized simulation data, can effectively classify reaction events in neutron-induced reactions on carbon, outperforming manual cut methods.

Contribution

It introduces a neural network-based classification approach for reaction data, using voxelized training datasets derived from Geant4 simulations, for the first time in this context.

Findings

Neural networks outperform manual cuts in classification accuracy.

Voxel-based training improves neural network performance.

Method proves feasible for energy-differential measurements at CERN.

Abstract

The paper explores the feasibility of using machine learning techniques, in particular neural networks, for classification of the experimental data from the joint $^\text{nat}$C(n,p) and $^\text{nat}$C(n,d) reaction cross section measurement from the neutron time of flight facility n_TOF at CERN. Each relevant $\Delta E$-$E$ pair of strips from two segmented silicon telescopes is treated separately and afforded its own dedicated neural network. An important part of the procedure is a careful preparation of training datasets, based on the raw data from Geant4 simulations. Instead of using these raw data for the training of neural networks, we divide a relevant 3-parameter space into discrete voxels, classify each voxel according to a particle/reaction type and submit these voxels to a training procedure. The classification capabilities of the structurally optimized and trained neural networks are found to be superior to those of the manually selected cuts.