Dortmund 2021 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

T: Fachverband Teilchenphysik

T 40: Experimental methods II

T 40.7: Talk

Tuesday, March 16, 2021, 17:30–17:45, To

Neural network background estimation for Higgs boson pairs decaying to bbbb final state — •Marta Czurylo and André Schöning — Physikalisches Institut Universität Heidelberg

Monte Carlo (MC) simulations and data-driven techniques are commonly used for the background estimation in ATLAS analyses. An advantage of data-driven methods over more traditional MC simulations is observed when MC is unreliable, for example, in case of QCD processes.

One of such techniques is background reweighting from control to signal regions which is firstly discussed in general. A novel approach uses neural network machinery for the reweighting. The idea and preliminary performance of the neural network reweighting are presented for the Vector Boson Fusion production of Higgs boson pairs decaying to bbbb final state. The studies are performed based on ATLAS datasets collected between 2016 and 2018 with total integrated luminosity of 126.7 fb⁻¹.