Finding physics signals with shower deconstruction

By Davison E. Soper, Michael Spannowsky. 2011

153 citations recorded by the source corpus.

We introduce shower deconstruction, a method to look for new physics in a hadronic environment. The method aims to be a full information approach using small jets. It assigns to each event a number {chi} that is an estimate of the ratio of the probability for a signal process to produce that event to the probability for a background process to produce that event. The analytic functions we derive to calculate these probabilities mimic what full event generators like Pythia or Herwig do and can be depicted in a diagrammatic way. As an example, we apply this method to a boosted Higgs boson produced in association with a Z boson and show that this method can be useful to discriminate this signal from the Z+jets background.

Subfields: Artificial Intelligence, Nuclear and High Energy Physics

View Paper: https://www.semanticscholar.org/paper/ade5ac3eb20d741b03508986b3768720e89fb604

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Nuclear and High Energy Physics

Citations in the corpus

Pileup per particle identification (2014, methodology): “In figure 9 (left) we show the mass resolution9 for jets with pT between 100 GeV and 200 GeV at nPU = 80.”
Jet substructure at the Tevatron and LHC: new results, new tools, new benchmarksReport prepared by the participants of the BOOST 2011 Workshop at Princeton University, 22–26 May 2011. L Asquith (lasquith@hep.anl.gov), S Rappoccio (rappocc@fnal.gov) and C K Vermilion (verm@uw.edu), editors. (2011, methodology): “The template overlap method [42] and the shower deconstruction method [43] classify jets with the help of approximations to hard matrix elements and the parton shower.”
Gaining (mutual) information about quark/gluon discrimination (2014, methodology): “[60], one could use the ratio of the quark/gluon double differential distributions as a weighting factor.”
Jet-images: computer vision inspired techniques for jet tagging (2014, background): “…and jet tagging / jet ﬂavor identiﬁcation algorithms [2–12], jet grooming and pileup mitigation [13–15], template and matrix element techniques [16, 17], energy ﬂow and event shapes [18–24], and probabilistic jet clustering [25, 26]; more inclusive summaries can be found in references [27–30]…”
Playing tag with ANN: boosted top identification with pattern recognition (2015, methodology): “More recently, Matrix Element Method [30, 31] inspired techniques such as Shower Deconstruction have emerged [32, 33], where a boosted jet is tagged using approximations to hard matrix elements and the parton shower.”
Efficient Probabilistic Inference in the Quest for Physics Beyond the Standard Model (2018, background): “Until now, inference for the latent structure of an individual event has only been possible by accepting a drastic simpliﬁcation of the high-ﬁdelity simulators [4, 7–10, 15, 23, 27, 37, 38, 46, 60, 67, 68, 79, 81].”
Maximizing boosted top identification by minimizing N-subjettiness (2011, background): “There has also been work on template and matrix element methods [31, 32].”
The Lund jet plane (2018, methodology): “will be evaluated on an independent sample of 200;000 signal and background jets. 5 It also highlights an important dierence between our log-likelihood approach and the showerdeconstruction approach [67]. Both are partially analytical multi-variate log-likelihood approaches. Showerdeconstruction exp”
Looking Inside Jets (2019, methodology): “BSTRUCTURE: CONCEPTS AND TOOLS 82 of all the weights associated with fpNgunder the signal or background hypothesis. To compute the weight for a given history, one uses a Feynman-diagrammatic approach [164, 165] where each vertex receives a factor of the form He R with Ha partonic splitting probabili”
JUNIPR: a framework for unsupervised machine learning in particle physics (2018, methodology): “lity density for those particular momenta to arise in an event, conditioned on the event selection criteria used to select the training data. As a complementary example of this, shower deconstruction [56,57] provides a theory-driven approach to probabilistic modeling in particle physics, in which pr”
Jet substructure at the Large Hadron Collider (2018, methodology): “(c) Shower Deconstruction [180, 181] was designed to be analogous to running a parton shower Monte Carlo generator in reverse, where emission and decay probabilities at each vertex, color connections,”