Detecting an axion-like particle with machine learning at the LHC

Axion-like particles (ALPs) appear in various new physics models with spontaneous global symmetry breaking. When the ALP mass is range of MeV to GeV, cosmology and astrophysics bounds are so far quite weak. In this work, we investigate such light ALPs through ALP-strahlung production processes $pp \to W^\pm a, Z a$ sequential decay $a \gamma\gamma$ at 14 TeV LHC an integrated luminosity 3000 fb$^{-1}$ (HL-LHC). Building on concept jet image which uses calorimeter towers as pixels measures a image, potential machine learning techniques based convolutional neural network (CNN) identify highly boosted pair collimated photons. With CNN tagging algorithm, demonstrate that our approach can extend current sensitivity probe from 0.3~GeV 5~GeV. The obtained stronger than existing limits ALP-photon coupling.