Enhancing the performance of flavor tagging algorithms in the ATLAS experiment with low-momentum tracks

Student: Madison Ambriz

Faculty Mentor: Alexandra Miller

Physics & Astronomy

College of Science, Technology, and Business

The Large Hadron Collider (LHC) is the world's largest and most powerful particle accelerator, where protons travel close to the speed of light before colliding with a center of mass energy of √s = 13.6 TeV. The ATLAS experiment records these proton-proton collisions, allowing us to probe the Standard Model of particle physics. Flavor-tagging is the identification of jets from b-quarks, c-quarks, and light-flavored quarks originating from the proton-proton collisions. Flavor-tagging is of major importance to the ATLAS experiment, particularly in identifying Higgs boson decays to bottom quarks and in searches for new phenomena involving heavy-flavor quarks. Flavor-tagging algorithms in ATLAS exploit the long lifetime, high mass, and high decay multiplicity of b- and c-hadrons, as well as the properties of heavy-quark fragmentation. They take charged particle tracks, primary vertices, and hadronic jets as inputs and train various high-level and low-level machine learning methods to identify heavy-flavor jets. At the 77% b-jet identification efficiency working point of the DL1r flavor-tagging algorithm, the light-jet rejection factor peaks at O(500) for jets with pT = 150 GeV but drops to less than 200 for jets with pT < 50 GeV [1]. This drop in the light-jet rejection factor may partly result from some decay products of the b- and c-hadrons not being properly reconstructed due to the minimum pT requirement of 500 MeV in the standard ATLAS track reconstruction. This poster will present a study to determine whether including charged particle tracks with pT < 500 MeV in the flavor-tagging algorithms can improve the flavor-tagging performance for low-pT jets. [1] https://arxiv.org/pdf/2211.16345. Work supported by the U.S. Department of Energy, Office of Science, Division of High Energy Physics, RENEW-Initiative, GROWTH-MSI Program #DE-SC0023725.