-
Robust Anomaly Detection for Particle Physics Using Multi-Background Representation Learning
Authors:
Abhijith Gandrakota,
Lily Zhang,
Aahlad Puli,
Kyle Cranmer,
Jennifer Ngadiuba,
Rajesh Ranganath,
Nhan Tran
Abstract:
Anomaly, or out-of-distribution, detection is a promising tool for aiding discoveries of new particles or processes in particle physics. In this work, we identify and address two overlooked opportunities to improve anomaly detection for high-energy physics. First, rather than train a generative model on the single most dominant background process, we build detection algorithms using representation…
▽ More
Anomaly, or out-of-distribution, detection is a promising tool for aiding discoveries of new particles or processes in particle physics. In this work, we identify and address two overlooked opportunities to improve anomaly detection for high-energy physics. First, rather than train a generative model on the single most dominant background process, we build detection algorithms using representation learning from multiple background types, thus taking advantage of more information to improve estimation of what is relevant for detection. Second, we generalize decorrelation to the multi-background setting, thus directly enforcing a more complete definition of robustness for anomaly detection. We demonstrate the benefit of the proposed robust multi-background anomaly detection algorithms on a high-dimensional dataset of particle decays at the Large Hadron Collider.
△ Less
Submitted 16 January, 2024;
originally announced January 2024.
-
Model selection and signal extraction using Gaussian Process regression
Authors:
Abhijith Gandrakota,
Amitabh Lath,
Alexandre V. Morozov,
Sindhu Murthy
Abstract:
We present a novel computational approach for extracting weak signals, whose exact location and width may be unknown, from complex background distributions with an arbitrary functional form. We focus on datasets that can be naturally presented as binned integer counts, demonstrating our approach on the CERN open dataset from the ATLAS collaboration at the Large Hadron Collider, which contains the…
▽ More
We present a novel computational approach for extracting weak signals, whose exact location and width may be unknown, from complex background distributions with an arbitrary functional form. We focus on datasets that can be naturally presented as binned integer counts, demonstrating our approach on the CERN open dataset from the ATLAS collaboration at the Large Hadron Collider, which contains the Higgs boson signature. Our approach is based on Gaussian Process (GP) regression - a powerful and flexible machine learning technique that allowed us to model the background without specifying its functional form explicitly, and to separate the background and signal contributions in a robust and reproducible manner. Unlike functional fits, our GP-regression-based approach does not need to be constantly updated as more data becomes available. We discuss how to select the GP kernel type, considering trade-offs between kernel complexity and its ability to capture the features of the background distribution. We show that our GP framework can be used to detect the Higgs boson resonance in the data with more statistical significance than a polynomial fit specifically tailored to the dataset. Finally, we use Markov Chain Monte Carlo (MCMC) sampling to confirm the statistical significance of the extracted Higgs signature.
△ Less
Submitted 11 February, 2022;
originally announced February 2022.
-
Searching for exotic B meson decays with the CMS L1 track trigger
Authors:
Jared A. Evans,
Abhijith Gandrakota,
Simon Knapen,
Hardik Routray
Abstract:
The CMS phase II track trigger may allow for a displaced dimuon vertex trigger with qualitatively lower $p_T$ thresholds than existing dimuon triggers. With this technique, we show that the CMS reach for exotic $B$-meson decays involving a displaced dimuon resonance, such as a light, Higgs-mixed scalar, can be competitive with that of LHCb and Belle II.
The CMS phase II track trigger may allow for a displaced dimuon vertex trigger with qualitatively lower $p_T$ thresholds than existing dimuon triggers. With this technique, we show that the CMS reach for exotic $B$-meson decays involving a displaced dimuon resonance, such as a light, Higgs-mixed scalar, can be competitive with that of LHCb and Belle II.
△ Less
Submitted 16 August, 2020;
originally announced August 2020.
-
Higgs Physics at the HL-LHC and HE-LHC
Authors:
M. Cepeda,
S. Gori,
P. Ilten,
M. Kado,
F. Riva,
R. Abdul Khalek,
A. Aboubrahim,
J. Alimena,
S. Alioli,
A. Alves,
C. Asawatangtrakuldee,
A. Azatov,
P. Azzi,
S. Bailey,
S. Banerjee,
E. L. Barberio,
D. Barducci,
G. Barone,
M. Bauer,
C. Bautista,
P. Bechtle,
K. Becker,
A. Benaglia,
M. Bengala,
N. Berger
, et al. (352 additional authors not shown)
Abstract:
The discovery of the Higgs boson in 2012, by the ATLAS and CMS experiments, was a success achieved with only a percent of the entire dataset foreseen for the LHC. It opened a landscape of possibilities in the study of Higgs boson properties, Electroweak Symmetry breaking and the Standard Model in general, as well as new avenues in probing new physics beyond the Standard Model. Six years after the…
▽ More
The discovery of the Higgs boson in 2012, by the ATLAS and CMS experiments, was a success achieved with only a percent of the entire dataset foreseen for the LHC. It opened a landscape of possibilities in the study of Higgs boson properties, Electroweak Symmetry breaking and the Standard Model in general, as well as new avenues in probing new physics beyond the Standard Model. Six years after the discovery, with a conspicuously larger dataset collected during LHC Run 2 at a 13 TeV centre-of-mass energy, the theory and experimental particle physics communities have started a meticulous exploration of the potential for precision measurements of its properties. This includes studies of Higgs boson production and decays processes, the search for rare decays and production modes, high energy observables, and searches for an extended electroweak symmetry breaking sector. This report summarises the potential reach and opportunities in Higgs physics during the High Luminosity phase of the LHC, with an expected dataset of pp collisions at 14 TeV, corresponding to an integrated luminosity of 3 ab$^{-1}$. These studies are performed in light of the most recent analyses from LHC collaborations and the latest theoretical developments. The potential of an LHC upgrade, colliding protons at a centre-of-mass energy of 27 TeV and producing a dataset corresponding to an integrated luminosity of 15 ab$^{-1}$, is also discussed.
△ Less
Submitted 19 March, 2019; v1 submitted 31 January, 2019;
originally announced February 2019.
