Bulletin of the American Physical Society
APS April Meeting 2022
Volume 67, Number 6
Saturday–Tuesday, April 9–12, 2022; New York
Session Q09: Data Analysis, AI and ML IIRecordings Available
|
Hide Abstracts |
Sponsoring Units: DPF GDS Chair: Matthew Bellis, Siena College Room: Salon 3 |
Monday, April 11, 2022 10:45AM - 10:57AM |
Q09.00001: Event-based anomaly detection for new physics searches at the LHC using machine learning Sergei Chekanov, Walter Hopkins This paper discusses model-agnostic searches for new physics at the Large Hadron Collider (LHC) using anomaly-detection techniques for the identification of event signatures that deviate from the Standard Model (SM). We investigate anomaly detection in the context of machine-learning approaches using autoencoders, and illustrate expected shapes of invariant masses in the outlier region using Monte Carlo simulations. Challenges and conceptual limitations of this approach are discussed. |
Monday, April 11, 2022 10:57AM - 11:09AM |
Q09.00002: Clustering for interpreting complex high-energy physics models Walter Hopkins, Evangelos Kourlitis After discovering the last piece of the Standard Model (SM), the Higgs boson, experiments at the Large Hadron Collider (LHC) have been searching for hints of physics Beyond the SM (BSM) to yield insights into these phenomena. These searches have not yet produced any significant deviations from SM predictions. Identifying unexplored regions in experimental observable space (object pTs, MET, etc) is essential in developing future BSM searches. One tool for finding new search regions is the Phenomenological Minimal Supersymmetric Standard Model (pMSSM) scan that is currently ongoing within the ATLAS collaboration. pMSSM models that have not been excluded by current searches can be used to build new search regions. Manually interpreting the high-dimensional space of our observables for each non-excluded model can however be challenging. Unsupervised data exploration algorithms (e.g. clustering) can analyze all of the non-excluded models and identify groups of non-excluded pMSSM models that live in a similar region of observable space. This could reduce thousands of theory models into a significantly smaller number of proto-search regions. These regions can then be developed into new BSM search regions. We present results from applying k-means clustering and dimensional reduction in the form of an autoencoder applied to well-understood simplified SUSY models. These results show that models can be grouped together in an unsupervised manner. |
Monday, April 11, 2022 11:09AM - 11:21AM |
Q09.00003: Neural Network Tagger for Single Higgs Boson that Decays into Two W Bosons Zichun Hao, Javier M Duarte, Raghav Kansal, Cristina M Suarez Although the experimental discovery of the Higgs boson by the Large Hadron Collider at CERN marked the completion of the standard model (SM), there are still many mysteries that the SM cannot explain and may require us to know more about the Higgs boson. Discriminating the Higgs boson from the background is the first step towards more precise measurements and better insights into the Higgs mechanism. One decay mode of the Higgs boson is the WW* mode, in which a Higgs boson decays into a pair of W bosons (HWW), which has been considered a difficult final state to analyze due to its low signal-to-background ratio. We adapt the ParticleNet deep graph neural network architecture to identify the HWW topology and tune the model and input selections. We have achieved so far an area under the receiver operating characteristic (ROC) curve (AUC) of 91.5% and the signal efficiency of about 40% at 1% background efficiency. Besides, the dependency of performance on the Higgs mass and transverse momentum has been largely reduced. |
Monday, April 11, 2022 11:21AM - 11:33AM |
Q09.00004: Searches for new physics in collision events using a statistical technique for anomaly detection Sergei Chekanov, Jacob E Crosby We discuss a statistical anomaly-detection method for model-independent searches for new physics in collision events produced at the Large |
Monday, April 11, 2022 11:33AM - 11:45AM |
Q09.00005: Interaction Network Autoencoder in the Level-1 Trigger Sukanya S Krishna In the LHC, the FPGA-based real-time data filter system that rapidly decides which collision events to record, known as the level-1 trigger, requires small models because of the low latency budget and other computing resource constraints. To enhance the sensitivity to unknown new physics, we want to put generic anomaly detection algorithms into the trigger. Past research suggests that graph neural network (GNN) based autoencoders can be effective mechanisms for reconstructing particle jets and isolating anomalous signals from background data. Rather than treating particle jets as ordered sequences or images, interaction networks embed particle jet showers as a graph and exploit particle-particle relationships to efficiently encode and reconstruct particle-level information within jets. This project investigates graph-based standard and variational autoencoders. The two objectives in this project are to evaluate the anomaly detection performance against other kinds of autoencoder structures (e.g. convolutional or fully-connected), and implement the model on an FPGA to meet L1 trigger requirements. |
Monday, April 11, 2022 11:45AM - 11:57AM |
Q09.00006: Point Cloud Deep Learning Methods for Pion Reconstruction in the ATLAS Detector Mariel Pettee Reconstructing the type and energy of isolated pions from the ATLAS calorimeters is a key step in hadronic reconstruction. The existing methods were optimized early in the experiment lifetime. We recently showed that image-based deep learning can significantly improve the performance over these traditional techniques. This note presents an extension of that work using point cloud methods that do not require calorimeter clusters to be projected onto a fixed and regular grid. Instead, we use transformer, deep sets, and graph neural network architectures to process calorimeter clusters as point clouds. We demonstrate the performance of these new approaches as an important step towards a full deep learning-based low-level hadronic reconstruction. |
Monday, April 11, 2022 11:57AM - 12:09PM |
Q09.00007: Calibration of electrons and photons in the CMS ECAL with graph neural networks Simon Rothman The Compact Muon Solenoid (CMS) detector is a general-purpose detector on the energy frontier of particle physics at the CERN Large Hadron Collider (LHC). Products of proton-proton collisions at a center-of-mass energy of 13 TeV are reconstructed in the CMS detector to probe the standard model of particle physics and to search for processes beyond the standard model. The development of precision algorithms for this reconstruction is therefore a key objective in optimizing the precision of all CMS physics results. While machine learning techniques are now prevalent at CMS for these tasks, they have largely relied on high-level human-engineered input features. However, much of the disruptive impact of machine learning in other fields has been realized by bypassing human feature engineering and instead training deep learning algorithms on low-level data. We have developed a novel machine learning architecture based on dynamic graph neural networks which allows regression directly on low-level detector hits, and we have applied this model to the calibration of electron and photon energies in CMS. In this work we will discuss our new architecture and show its performance in predicting the electron and photon energies used in physics analyses at CMS. |
Monday, April 11, 2022 12:09PM - 12:21PM |
Q09.00008: EMD Neural Network for HGCAL Data Compression Encoder ASIC Rohan Shenoy The High Granularity Calorimeter (HGCAL) is part of the High Luminosity upgrade of the CMS detector at the Large Hadron Collider (HL-LHC). For the trigger primitive generation of the 6 million channels in this detector, data compression at the front end may be accomplished by using deep-learning techniques using an on-ASICs network. The ASIC foresees an encoder based on a convolutional neural network (CNN). The performance is evaluated using the energy mover's distance (EMD). Ideally, we would like to quantify the loss between the input and the decoded image at every step of the training using the EMD. However, the EMD is not differentiable and can therefore not be used directly as a loss function for gradient descent. The task of this project is to approximate this EMD using a separate set of CNNs and then implement the EMD NN as a custom loss for the ASIC encoder training, with the goal of achieving better physics performance. |
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. |
© 2024 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