Bulletin of the American Physical Society
2024 APS April Meeting
Wednesday–Saturday, April 3–6, 2024; Sacramento & Virtual
Session F14: Mini-Symposium: ML/AI/Computation on the Front EndMini-Symposium
|
Hide Abstracts |
Sponsoring Units: DPF DNP Chair: Dmitri Denisov, Brookhaven National Laboratory Room: SAFE Credit Union Convention Center Ballroom B3, Floor 2 |
Thursday, April 4, 2024 8:30AM - 9:06AM |
F14.00001: Artificial Intelligence and Machine Learning in Particle Physics Invited Speaker: Michael Kagan Artifical Intelligence and Machine Learning (AI/ML) have become integral parts of particle physics data analysis pipelines. This talk will give an overview of AI/ML in particle physics and will discuss recent progress and future directions for AI/ML research and development for particle physics. |
Thursday, April 4, 2024 9:06AM - 9:18AM |
F14.00002: A Model-Independent Boosted Decision Tree Approach to Event Classification in MicroBooNE Daniel Xing MicroBooNE is a 170-ton Liquid argon time projection chamber (LArTPC) detector located in the Booster Neutrino Beam (BNB) beamline at Fermilab. One of the primary goals of MicroBooNE is investigating the anomalous excess of low energy electron neutrino like signals observed by other short-baseline experiments over the past two decades. With its LArTPC technology, MicroBooNE can probe and distinguish electromagnetic showers originating from both electron neutrino interactions as well as background photons, a vast improvement over prior experiments. Recent MicroBooNE efforts have focused on specific signal channels, looking for signals originating from one model or a particular class of models. In this talk, we present novel work towards the next generation photon and electron-positron separation techniques, using a modern GPU enhanced Boosted Decision Tree (BDT) framework utilizing XGBoost (Extreme Gradient Boosting) and other powerful machine learning libraries to train a single model-independent BDT to separate out various important background and signal categories. |
Thursday, April 4, 2024 9:18AM - 9:30AM |
F14.00003: DeepCore 2.0: Convolutional Neural Network for Tracking in Jets with High Transverse Momentum Hichem Bouchamaoui, Nicholas J Haubrich, Soohyun Yoon, James D Olsen Tracking of charged particles in dense environments, especially in the core of high transverse-momentum ($p_T$) jets, presents a growing challenge with increasing LHC luminosity. Despite the CMS phase-1 pixel detector upgrade, and dedicated cluster splitting and pattern recognition algorithms like JetCore, there is still significant room for improvement. Limiting the computation time for track reconstruction represents an additional challenge as the number of proton-proton interactions per crossing (pileup) increases. DeepCore is a machine learning algorithm designed to improve track seeding in the core of high-$p_T$ jets in the presence of increased pileup. In this talk, we summarize recent improvements to DeepCore optimized in the context of a hybrid with JetCore, leading to a significant increase in track reconstruction efficiency relative to JetCore alone for particles with $p_T$ above 10 GeV. This improved algorithm, referred to as DeepCore2.0, also leads to a reduction in overall computation time for track reconstruction, with further reduction possible in the future. |
Thursday, April 4, 2024 9:30AM - 9:42AM |
F14.00004: Anomaly Detection: Tips, Tricks and Examples on How to Find Outlying Events in the Standard Model Using Machine Learning Jacob E Crosby How many phase spaces in the Standard Model have not been analyzed, let alone thought of? In the past few years, interest has grown rapidly in using machine learning for anomaly detection in High Energy Physics in both the ATLAS and CMS experiment located at CERN Large Hadron Collider. This approach helps define anomalous phase spaces within the Standard Model (SM) that may contain Beyond the Standard Model (BSM) events. Several approaches have been developed and successfully applied to find these anomalous phase spaces. How are these techniques developed? What metrics can be used to define an effective phase space? Can these techniques increase discovery sensitivity? In this presentation, I will explore various BSM model agnostic approaches, discuss results using example BSM models such as the radion and dark-matter, and offer practical insights on formulating an Anomaly Detection technique. |
Thursday, April 4, 2024 9:42AM - 9:54AM |
F14.00005: Smart Pixels: Algorithm design and hardware testing for a 28m ROIC for future pixel trackers Carissa N Kumar, Emily Pan, Karri F DiPetrillo, Anthony Badea, Jennet Dickinson, Jieun Yoo, Morris L Swartz, Giuseppe Di Guglielmo, Alice L Bean, Douglas R Berry, Manuel Blanco Valentín, Farah Fahim, Lindsey A Gray, James F Hirschauer, Shruti Kulkarni, Ronald J Lipton, Petar Maksimovic, Corrinne Mills, Benjamin Parpillon, Gauri Pradhan, Nhan V Tran, Aaron Young, Chinar Syal, Dahai Wen Disentangling the enormous number of particles produced at high energy colliders calls for cutting-edge silicon pixel detectors. These tracking detectors reconstruct the paths of charged particles, an essential experimental task. With billions of readout channels and event rates as high as 40 MHz, these detectors will generate petabytes of data per second. New technologies are needed for ultrafast and power-efficient data extraction. We show here work to design a readout integrated circuit (ROIC) with an on-chip machine learning (ML) algorithm to perform data reduction at the source. This work highlights the algorithm and hardware co-design, giving insight for a future 28nm ROIC implementation. |
Thursday, April 4, 2024 9:54AM - 10:06AM |
F14.00006: Boosted and Resolved Jet Assignment using Symmetry-Preserving Attention Networks in Nonresonant Multi-Higgs-Boson Events Haoyang Li, Marko Stamenkovic, Alexander K Shmakov, Michael A Fenton, Caden Mikkelsen, Jovan Mitic, Cristina Mantilla Suarez, Melissa K Quinnan, Harvey B Newman, Pierre Baldi, Daniel Whiteson, Javier M Duarte The Higgs boson’s self-coupling has a significant impact on the production rate of multiple Higgs bosons. Measuring the self-coupling at the CERN LHC is crucial because any deviations from our expectations could potentially lead to new discoveries of physics beyond the standard model of particle physics. Most events are fully hadronic, meaning every Higgs boson decays to a bottom quark-antiquark pair. This introduces a combinatorial challenge known as the jet assignment problem, in which jets are assigned to Higgs boson candidates. For a given event topology, symmetry-preserving attention networks (SPA-Nets) have been introduced to address this challenge. However, the complexity of this challenge increases when considering different reconstruction topologies for each Higgs boson candidate simultaneously, i.e., two “resolved'' small-radius jets each containing a cascade initiated by a bottom quark or one “boosted'' large-radius jet containing a merged cascade initiated by a bottom quark-antiquark pair. In this work, we generalize the SPA-Net approach to simultaneously consider both boosted and resolved reconstruction possibilities and unambiguously interpret an event as “fully resolved,'' “fully boosted,'' or in between. We report the performance of baseline methods, the original SPA-Net approach, and our generalized version on nonresonant HH and HHH production simulated by Pythia and Delphes. |
Thursday, April 4, 2024 10:06AM - 10:18AM |
F14.00007: Abstract Withdrawn |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2025 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700