Powered by Indico
v3.2.3
Abstract:
One of the LHC’s main goals after the Higgs discovery is to measure the Higgs self-coupling $\lambda_{3H}$ through the rare production of Higgs boson pairs. Existing searches rely on cut-based methods or boosted decision trees (BDTs), but these approaches face intrinsic limitations.
We propose an alternative based on the Matrix Element Method (MEM), a statistically optimal multivariate technique that directly incorporates theoretical amplitudes and detector effects. MEM has historically enabled breakthroughs in rare process measurements (e.g. the top-quark mass at Tevatron, single-top production at the LHC) and is particularly well suited to probing rare di-Higgs production.
While most MEM applications have been limited to leading-order (LO) accuracy, we developed the first NLO implementation of MEM for the process $gg \to HH \to b\bar b \gamma\gamma$. This framework, built within an extended version of MoMEMta, incorporates NLO matrix elements from POWHEG-BOX-V2, and can be generalized to SMEFT. We demonstrate its functionality and competitive sensitivity using Monte Carlo samples. This seminar will be the opportunity to present our main results, positioning MEM@NLO as a promising alternative for $\lambda_{3H}$ extraction at the HL-LH.
In parallel, we will also present our contributions to the ATLAS software developments, by working on the ACTS interface within Athena, enabling modernized track reconstruction and smoother integration of advanced tools into the ATLAS reconstruction framework.
Zoom direct link:
https://cern.zoom.us/j/4042031778?pwd=N0RlZ0l3dGtEYk9zT25OWmlwVklSZz09
Alice Ling Lin