Particle identification with the Belle II calorimeter using machine learning

TL;DR

This paper demonstrates that a convolutional neural network can effectively distinguish muons from pions in the Belle II calorimeter, enhancing particle identification accuracy at low momenta using energy deposition images.

Contribution

The study introduces a CNN-based approach for particle identification in the Belle II calorimeter, improving separation of muons and pions over traditional methods.

Findings

CNN outperforms existing PID methods in low-momentum regions

Energy deposition shape differences are effectively exploited by the CNN

Simulation results confirm improved muon-pion separation

Abstract

I present an application of a convolutional neural network (CNN) to separate muons and pions in the Belle II electromagnetic calorimeter (ECL). The ECL is designed to measure the energy deposited by charged and neutral particles. It also provides important contributions to the particle identification (PID) system. Identification of low-momenta muons and pions in the ECL is crucial if they do not reach the outer muon detector. Track-seeded cluster energy images provide the maximal possible information. The shape of the energy depositions for muons and pions in the crystals around an extrapolated track at the entering point of the ECL is used together with crystal positions in $\theta-\phi$ plane and transverse momentum of the track to train a CNN. The CNN exploits the difference between the dispersed energy depositions from pion hadronic interactions and the more localized muon electromagnetic interactions. Using simulation, the performance of the CNN algorithm is compared with other PID methods at Belle II which are based on track-matched clustering information. The results show that the CNN PID method improves muon-pion separation in low momentum.