Probing triple Higgs coupling with machine learning at the LHC

TL;DR

This paper explores using machine learning, specifically Message Passing Neural Networks, to improve the measurement of the triple Higgs coupling at the LHC, providing bounds on the coupling's possible values.

Contribution

It introduces a novel application of MPNN to Higgs pair production analysis, enhancing signal significance and setting new bounds on the triple Higgs coupling.

Findings

MPNN improves signal significance in Higgs pair production

A 2σ upper bound on the Higgs pair production cross section is 3.7 times the SM prediction

Limits the triple Higgs coupling to between -3 and 11.5

Abstract

Measuring the triple Higgs coupling is a crucial task in the LHC and future collider experiments. We apply the Message Passing Neural Network (MPNN) to the study of the non-resonant Higgs pair production process $pp \to hh$ in the final state with $2b + 2\ell + E_{\rm T}^{\rm miss}$ at the LHC. Although the MPNN can improve the signal significance, it is still challenging to observe such a process at the LHC. We find that a $2\sigma$ upper bound (including a 10\% systematic uncertainty) on the production cross section of the Higgs pair is 3.7 times the predicted SM cross section at the LHC with the luminosity of 3000 fb$^{-1}$, which will limit the triple Higgs coupling to the range of $[-3,11.5]$.