Probing the dimuon channel of a Z ′ Z^{\prime} boson at the HL-LHC using multivariate analysis

Ali Muhammad H. H, El-sayed A. El-dahshan, S. Elgammal · 2025

Read Probing the dimuon channel of a Z ′ Z^{\prime} boson at the HL-LHC using multivariate analysis by Ali Muhammad H. H, El-sayed A. El-dahshan, S. Elgammal (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-02739.txt.

Introduction

A dark Higgs signal at the future High-Luminosity LHC is hunted by Ali Muhammad and co-authors in a 2025 paper. The upgrade will raise integrated luminosity a lot, which sharpens Standard Model measures and improves sensitivity to rare events and new physics. The target here is the leptonic decay of a Z prime boson, in the dimuon final state.

A multivariate analysis separates the dark Higgs signal from the dominant Standard Model background. The study uses the Toolkit for Multivariate Analysis and compares three classifiers: boosted decision trees, deep neural networks, and likelihood estimators. Conditions match the high-luminosity runs at fourteen TeV. Readers get a comparison of classifier performance for a search where the signal is rare and the background is large.

Catalog record

CREATORAli Muhammad H. H, El-sayed A. El-dahshan, S. Elgammal
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE27 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-02739

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Probing the dimuon channel of a Z ′ Z^{\prime} boson at the HL-LHC using multivariate analysis Ali Muhammad H. H, El-sayed A. El-dahshan, S. Elgammal Abstract The upcoming upgrade of the existing LHC facility at CERN is known as the High-Luminosity Large Hadron Collider (HL-LHC). It is designed to extend its physics reach by substantially increasing the integrated luminosity. It will enable more precise measurements of the Standard Model (SM) and improve sensitivity to rare events and possible new physics signatures. This study adopts a multivariate analysis (MVA) approach to effectively discriminate the dark Higgs (DH) signal against the dominant SM background. The analysis targets the leptonic decay mode of the Z ′ Z^{\prime} boson, focusing on the dimuon final state at s = 14 , TeV \sqrt{s}=14,\text{TeV} and 3000 , fb − 1 3000,\text{fb}^{-1} integrated luminosity corresponding to the HL-LHC. The DH signal is examined using the Toolkit for Multivariate Analysis (TMVA), employing and comparing the performance of three classifiers: Boosted Decision Trees (BDT), Deep Neural Networks (DNN), and Likelihood estimators. Keywords Dark Matter, Dark Higgs, Multivariate Analysis, The HL-LHC I Introduction Dark matter (DM) is a hypothetical form of matter that does not interact electromagnetically, making it invisible to all current forms of electromagnetic observation Bertone et al. 2005 ; Feng 2010 . Evidence for its presence arises from a variety of astrophysical and cosmological measurements, such as galaxy rotation curves Rubin et al. 1980 ; Sofue and Rubin 2001 , gravitational lensing phenomena Clowe et al. 2006 ; Kneib and Natarajan 2011 , and fluctuations observed in the cosmic microwave background Planck Collaboration 2020 ; Hinshaw et al. 2013 . These observations collectively suggest that DM constitutes about 27% of the total energy density of the universe Planck Collaboration 2020 ; Particle Data Group 2022 . Despite its gravitational influence, its fundamental particle properties remain elusive. Many potential candidates have been suggested, such Weakly Interacting Massive Particles (WIMPs) Jungman et al. 1996 ; Arcadi et al. 2020 ; Mahmoud et al. 2025 , axions Peccei and Quinn 1977 ; Sikivie 1983 , sterile neutrinos Dodelson and Widrow 1994 ; Abazajian et al. 2001 , and particles from dark sectors such as dark photons or Z^{\prime} bosons Alexander et al. 2016 ; Holdom 1986 . Various experimental strategies are employed to search for DM, including direct detection through nuclear recoil measurements Aprile et al. 2018 ; Cui et al. 2017 , indirect searches based on astrophysical observations Gaskins 2016 , and collider-based approaches targeting missing transverse energy signatures Boveia and Doglioni 2018 ; Aaboud et al. 2017 . Although the Standard Model (SM) has achieved great success, most recently with the discovery of the Higgs boson by the ATLAS and CMS collaborations in 2012 CMS Collaboration 2012 ; ATLAS Collaboration 2012 , it nevertheless fails to resolve fundamental issues, such as uncovering the particle nature of DM. Therefore, physicists have explored beyond-the-Standard-Model (BSM) scenarios. One such framework is the Dark Higgs (DH) model with a mono- Z^{\prime} portal Autran et al. 2015 . In this framework, DM particles are generated in proton–proton (pp) collisions via a heavy neutral mediator, the Z^{\prime} boson. This scenario has been studied in both Run-I Elgammal et al. 2023 and Run-II ATLAS Collaboration 2023 of the LHC. This work examines the DH model at the High-Luminosity phase of the LHC (HL-LHC), which is the upcoming upgrade of the LHC, operating at 14\,\mathrm{TeV} center-of-mass energy ( \sqrt{s} ), and a high integrated luminosity ( \mathcal{L} ) corresponding to 3000\,\mathrm{fb}^{-1} Apollinari et al. 2017 ; CERN 2020 . Signal and SM background samples are generated using Monte Carlo (MC) simulations. To distinguish the signal from the dominant SM backgrounds, the work employs multivariate analysis (MVA) using the TMVA package built in the ROOT framework. This approach has been applied in previous works CMS Collaboration 2025 ; Choudhury et al. 2024 ; Bhardwaj et al. 2020 ; Rana and Dubey 2024 and in recent HL-LHC studies Baradia et al. 2024 ; Chang et al. 2020 ; Adhikary et al. 2020 . We compare three classifiers: Boosted Decision Trees (BDT), Deep Neural Networks (DNN), and Likelihood estimators, which were used in previous studies, such as Baradia et al. 2024 ; Chang et al. 2020 ; Adhikary et al. 2020 ; Roe et al. 2005 ; Baldi et al. 2014 ; Cowan et al. 2011 ; Yalcin Kuzu 2023 . The structure of the paper is as follows. Section II briefly introduces the simplified DH model. Section III describes the HL-LHC. Section IV presents the techniques used for event generation in the MC simulations of the DH signals and its corresponding SM backgrounds. The TMVA package is introduced in Section V , and the MVA and its results are discussed in Section VI . Finally, Section VII provides a summary and conclusion. II The Simplified Model in the Mono- Z^{\prime} Portal The study in Ref. Autran et al. 2015 proposed the potential generation of DM particles associated with a neutral heavy gauge boson, denoted as Z^{\prime} , through the fusion of q\bar{q} pairs in pp collisions. In this framework, the Z^{\prime} mediates the interaction between SM particles and dark sector particles, with coupling constants g_{SM} and g_{DM} , respectively. Three simplified models were introduced in Autran et al. 2015 : the light vector (LV), the LV with inelastic effective field theory (EFT) coupling, and the DH models, with the DH model being the main focus of this paper. In the DH model, the Z^{\prime} boson mediates the production of a dark-sector Higgs boson ( h_{D} ), which subsequently decays into a pair of DM particles ( \chi\bar{\chi} ), following the heavy dark-sector (HDS) mass assumptions summarized in Table 1 . The Feynman diagram illustrating the DH model is shown in Fig. 1 . This study adopted the HDS mass assumption where M_{Z^{\prime}}=M_{h_{D}} for M_{Z^{\prime}}>125\,\mathrm{GeV} and takes the muonic decay of the Z^{\prime} , into dimuon, plus E_{T}^{\mathrm{miss}} , referring to DM particles, as the final signature of the interaction. The free parameters of the model are M_{Z^{\prime}} , M_{h_{D}} , g_{SM} and g_{DM} . The scope of this study for M_{Z^{\prime}} , and M_{h_{D}} ranges from 200 to 1000 \mathrm{GeV} , choosing g_{SM}=0.25 and g_{DM}=1.0 as recommended by ATLAS Collaboration 2023 to get the highest cross sections multiplied by the branching ratios ( \sigma\times\mathrm{Br} ) for the signal. Also, the range of 0.02 to 0.2 for g_{SM} was excluded for mass ranges from 200 to 1000 \mathrm{GeV} by that study ATLAS Collaboration 2023 performed by ATLAS. The studies carried out in Ref. Autran et al. 2015 ; Elgammal et al. 2023 found that \sigma\times\mathrm{Br} does not depend on the choice of the DM mass but on M_{Z^{\prime}} and M_{h_{D}} . This work aims to probe the DH model at the HL-LHC, corresponding to \sqrt{s}=14,\text{TeV} and \mathcal{L}=3000,\text{fb}^{-1} , using MC simulated samples generated as described in Section IV with g_{SM}=0.25 and g_{DM}=1.0 and M_{Z^{\prime}}=M_{h_{D}}\in[200,1000]\,\mathrm{GeV} . Figure 1 : Feynman diagram of the DH model, adapted from Ref. Autran et al. 2015 Scenario Mass assumptions Heavy dark sector M_{h_{D}}=\begin{cases}125\,\mathrm{GeV},&M_{Z^{\prime}}<125\,\mathrm{GeV},\\ M_{Z^{\prime}},&M_{Z^{\prime}}>125\,\mathrm{GeV}.\end{cases} Table 1 : Mass assumptions of the HDS for the particles produced in the DH model as outlined in Ref.

Read the complete text