A Deep Learning Approach to Particle Identification for the AMS Electromagnetic Calorimeter

Raheem Hashmani, Bilge Demirkoz, Emre Akbaş, Zhili Weng

Abstract

This talk presents a deep learning approach to particle identification for the Alpha Magnetic Spectrometer (AMS) electromagnetic calorimeter (ECAL). It focuses on rejecting the proton background in the cosmic-ray positron analysis by training deep learning models to distinguish electromagnetic showers from hadronic ones in the ECAL, improving background rejection over traditional methods. The corresponding study is published in Machine Learning: Science and Technology.

Date

Nov 4, 2022

Event

Fast Machine Learning for Science Workshop 2022

This talk was presented at the Fast Machine Learning for Science Workshop 2022. The corresponding study is published as “A Comparison of Deep Learning Models for Proton Background Rejection with the AMS Electromagnetic Calorimeter” in Machine Learning: Science and Technology.