Probing Electroweak Phase Transition in Extended Singlet Scalar Model with Resonant $HH$ production in $bbZZ$ Channel using Parameterized Machine Learning

TL;DR

This paper investigates the potential of detecting a heavy Higgs boson linked to a strong electroweak phase transition at the 14 TeV HL-LHC, utilizing a novel parameterized machine learning approach to improve signal-background discrimination.

Contribution

It introduces a parameterized machine learning method for resonant di-Higgs analysis in the $bbZZ$ channel within a singlet scalar extension of the Standard Model, focusing on electroweak phase transition signatures.

Findings

Potential to discover resonant di-Higgs signals up to 490 GeV.

Enhanced discrimination between signal and background using parameterized ML.

Feasibility of probing electroweak phase transition with HL-LHC data.

Abstract

In this paper, a collider signature of a heavy Higgs boson at $14$ TeV HL-LHC is studied, where the heavy Higgs boson decays into a pair of standard model Higgs boson, which further decays to $bbZZ$ state and subsequently to $bb\ell^{+} \ell^{-}\nu_{\ell} \nu_{\ell}$ final state. To study this, we consider singlet scalar extension of the standard model and select the parameter space and mass of the heavy Higgs boson such that it prefers a strong first-order electroweak phase transition. The study is done following the $bbZZ$ analysis of CMS Collaboration and further using parameterized machine learning for final discrimination which simplifies the training process along with an improved discrimination between signal and background over the range of benchmark points. Despite the lower branching fraction, this channel can be a potential probe of the electroweak phase transition with the data sets collected by the CMS and ATLAS experiments at the $14$ TeV HL-LHC with $3$ $\rm{ab}^{-1}$ of integrated luminosity and a production of resonant di-Higgs signal can be potentially discovered up to 490 GeV of resonance mass.