Stop(Zh) Neural Networks! A case study in 3rd generation SUSY searches using Machine Learning

by Gabriele D'Amen (Brookhaven National Laboratory (US))

Thursday 11 Jun 2020, 15:00 → 16:30 US/Eastern

The widespread use of Machine Learning techniques in the high energy physics community is driving the creation of tools to simplify the implementation of Neural Network methods in a wide variety of analyses. We have developed an artificial neural network framework (HTag) to identify bb pairs coming from the decay of Higgs bosons in events with large jet multiplicities. The use of HTag is expected to improve reconstruction efficiency in a wide kinematic region and to reduce invariant mass bias with respect to traditional reconstruction methods. Information reconstructed with HTag has been used in a search for direct top squark pair production in events with missing transverse momentum plus either a pair of jets consistent with the Standard Model Higgs boson decay to b-quarks or a same-flavour opposite-sign dilepton pair with invariant mass consistent with a Z boson. The analysis is performed using the proton-proton collision data at √s=13 TeV collected by the ATLAS experiment during the LHC Run-2, corresponding to an integrated luminosity of 139 fb−1. No excess is observed in the data with respect to the Standard Model predictions.

 

Zoom information:

 https://cern.zoom.us/j/97420636116?pwd=N083bERNQ3d4RTBFS0hXU1M5Vytvdz09

Meeting ID  974 2063 6116

Meeting Password  040606

GDamen_StopZhNN.pdf