skip to main content


Title: The Dark Machines Anomaly Score Challenge: Benchmark Data and Model Independent Event Classification for the Large Hadron Collider
We describe the outcome of a data challenge conducted as part of the Dark Machines (https://www.darkmachines.org) initiative and the Les Houches 2019 workshop on Physics at TeV colliders. The challenged aims to detect signals of new physics at the Large Hadron Collider (LHC) using unsupervised machine learning algorithms. First, we propose how an anomaly score could be implemented to define model-independent signal regions in LHC searches. We define and describe a large benchmark dataset, consisting of >1 billion simulated LHC events corresponding to 10\, fb^{-1} 10 f b − 1 of proton-proton collisions at a center-of-mass energy of 13 TeV. We then review a wide range of anomaly detection and density estimation algorithms, developed in the context of the data challenge, and we measure their performance in a set of realistic analysis environments. We draw a number of useful conclusions that will aid the development of unsupervised new physics searches during the third run of the LHC, and provide our benchmark dataset for future studies at https://www.phenoMLdata.org. Code to reproduce the analysis is provided at https://github.com/bostdiek/DarkMachines-UnsupervisedChallenge.  more » « less
Award ID(s):
2019786
NSF-PAR ID:
10323045
Author(s) / Creator(s):
; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; more » ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; « less
Date Published:
Journal Name:
SciPost Physics
Volume:
12
Issue:
1
ISSN:
2542-4653
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. A<sc>bstract</sc>

    Three searches are presented for signatures of physics beyond the standard model (SM) inττfinal states in proton-proton collisions at the LHC, using a data sample collected with the CMS detector at$$ \sqrt{s} $$s= 13 TeV, corresponding to an integrated luminosity of 138 fb1. Upper limits at 95% confidence level (CL) are set on the products of the branching fraction for the decay intoτleptons and the cross sections for the production of a new bosonϕ, in addition to the H(125) boson, via gluon fusion (ggϕ) or in association with b quarks, ranging from$$ \mathcal{O} $$O(10 pb) for a mass of 60 GeV to 0.3 fb for a mass of 3.5 TeV each. The data reveal two excesses for ggϕproduction with localp-values equivalent to about three standard deviations atmϕ= 0.1 and 1.2 TeV. In a search fort-channel exchange of a vector leptoquark U1, 95% CL upper limits are set on the dimensionless U1leptoquark coupling to quarks andτleptons ranging from 1 for a mass of 1 TeV to 6 for a mass of 5 TeV, depending on the scenario. In the interpretations of the$$ {M}_{\textrm{h}}^{125} $$Mh125and$$ {M}_{\textrm{h},\textrm{EFT}}^{125} $$Mh,EFT125minimal supersymmetric SM benchmark scenarios, additional Higgs bosons with masses below 350 GeV are excluded at 95% CL.

     
    more » « less
  2. A bstract The production of dark matter in association with Higgs bosons is predicted in several extensions of the Standard Model. An exploration of such scenarios is presented, considering final states with missing transverse momentum and b -tagged jets consistent with a Higgs boson. The analysis uses proton-proton collision data at a centre-of-mass energy of 13 TeV recorded by the ATLAS experiment at the LHC during Run 2, amounting to an integrated luminosity of 139 fb − 1 . The analysis, when compared with previous searches, benefits from a larger dataset, but also has further improvements providing sensitivity to a wider spectrum of signal scenarios. These improvements include both an optimised event selection and advances in the object identification, such as the use of the likelihood-based significance of the missing transverse momentum and variable-radius track-jets. No significant deviation from Standard Model expectations is observed. Limits are set, at 95% confidence level, in two benchmark models with two Higgs doublets extended by either a heavy vector boson Z ′ or a pseudoscalar singlet a and which both provide a dark matter candidate χ . In the case of the two-Higgs-doublet model with an additional vector boson Z ′, the observed limits extend up to a Z ′ mass of 3 TeV for a mass of 100 GeV for the dark matter candidate. The two-Higgs-doublet model with a dark matter particle mass of 10 GeV and an additional pseudoscalar a is excluded for masses of the a up to 520 GeV and 240 GeV for tan β = 1 and tan β = 10 respectively. Limits on the visible cross-sections are set and range from to 0.05 fb to 3.26 fb, depending on the missing transverse momentum and b -quark jet multiplicity requirements. 
    more » « less
  3. Kasieczka, Gregor ; Nachman, Benjamin ; Shih, David (Ed.)
    A new paradigm for data-driven, model-agnostic new physics searches at colliders is emerging, and aims to leverage recent breakthroughs in anomaly detection and machine learning. In order to develop and benchmark new anomaly detection methods within this framework, it is essential to have standard datasets. To this end, we have created the LHC Olympics 2020, a community challenge accompanied by a set of simulated collider events. Participants in these Olympics have developed their methods using an R&D dataset and then tested them on black boxes: datasets with an unknown anomaly (or not). This paper will review the LHC Olympics 2020 challenge, including an overview of the competition, a description of methods deployed in the competition, lessons learned from the experience, and implications for data analyses with future datasets as well as future colliders. 
    more » « less
  4. Abstract The production of heavy neutral mass resonances, $$\text {Z}^{\prime }$$ Z ′ , has been widely studied theoretically and experimentally. Although the nature, mass, couplings, and associated quantum numbers of this hypothetical particle are yet to be determined, current LHC experimental results have set strong constraints assuming the simplest beyond Standard Model (SM) hypotheses. We present a new feasibility study on the production of a $$\text {Z}^{\prime }$$ Z ′ boson at the LHC, with family non-universal couplings, considering proton–proton collisions at $$\sqrt{s} = 13$$ s = 13 and 14 TeV. Such a hypothesis is well motivated theoretically and it can explain observed differences between SM predictions and experimental results, as well as being a useful tool to further probe recent results in searches for new physics considering non-universal fermion couplings. We work under two simplified phenomenological frameworks where the $$\textrm{Z}^{\prime }$$ Z ′ masses and couplings to the SM particles are free parameters, and consider final states of the $$\text {Z}^{\prime }$$ Z ′ decaying to a pair of $$\textrm{b}$$ b quarks. The analysis is performed using machine learning techniques to maximize the sensitivity. Despite being a well motivated physics case in its own merit, such scenarios have not been fully considered in ongoing searches at the LHC. We note the proposed search methodology can be a key mode for discovery over a large mass range, including low masses, traditionally considered difficult due to experimental constrains. In addition, the proposed search is complementary to existing strategies. 
    more » « less
  5. null (Ed.)
    A bstract A search for new physics with non-resonant signals in dielectron and dimuon final states in the mass range above 2 TeV is presented. This is the first search for non-resonant signals in dilepton final states at the LHC to use a background estimate from the data. The data, corresponding to an integrated luminosity of 139 fb − 1 , were recorded by the ATLAS experiment in proton-proton collisions at a center-of-mass energy of $$ \sqrt{s} $$ s = 13 TeV during Run 2 of the Large Hadron Collider. The benchmark signal signature is a two-quark and two-lepton contact interaction, which would enhance the dilepton event rate at the TeV mass scale. To model the contribution from background processes a functional form is fit to the dilepton invariant-mass spectra in data in a mass region below the region of interest. It is then extrapolated to a high-mass signal region to obtain the expected background there. No significant deviation from the expected background is observed in the data. Upper limits at 95% CL on the number of events and the visible cross-section times branching fraction for processes involving new physics are provided. Observed (expected) 95% CL lower limits on the contact interaction energy scale reach 35.8 (37.6) TeV. 
    more » « less