Machine Learning approach to boosting neutral particles identifcation in the LHCb calorimeter

A Boldyrev,F Ratnikov,V Chekalina

doi:10.1088/1742-6596/1525/1/012096

Abstract

We present a new approach to identifcation of boosted neutral particles using Electromagnetic Calorimeter (ECAL) of the LHCb detector. The identifcation of photons and neutral pions is currently based on the geometric parameters which characterise the expected shape of energy deposition in the calorimeter. This allows to distinguish single photons in the electromagnetic calorimeter from overlapping photons produced from high momentum π0 decays. The novel approach proposed here is based on applying machine learning techniques to primary calorimeter information, that are energies collected in individual cells around the energy cluster. This method allows to improve separation performance of photons and neutral pions and has no signifcant energy dependence.

Highlights

GEM detector has been invented by F
Hit digitization The performance of a triple-GEM detector depends on the electronics used to readout the signal
A dedicated software to reconstruct and analyze has been developed to measure the performance of triple-GEM detectors

Summary

Introduction

GEM detector has been invented by F. To reconstruct and analyse the collected data a dedicated software has been developed. Extract the useful information from data such as electronic channel, charge and arrival time of the signal;. Different settings have been used to test the triple-GEM performance: different gas mixtures, various electrical field settings inside the detector, magnetic field scan, several angle configuration with respect to the beam, etc. 2. Hit digitization The performance of a triple-GEM detector depends on the electronics used to readout the signal. APV-25 is a commercial chip that samples the charge up to about 50 fC each 25 ns for 27 times after the trigger signal. While TIGER can return directly both the charge and time measurements, the signal from the APV-25 needs to be analyzed. The hit time thit is evaluated with the fit in the middle of the rising edge

Hit clusterization

Findings

Conclusion