Resolving Combinatorial Ambiguities in Dilepton tt ̄ Event Topologies with Neural Networks

by Mr Zhongtian Dong (University of Kansas, USA)

Thursday 10 Mar 2022, 09:45 → 11:45 Asia/Shanghai

Tsung-Dao Lee Institute (Meeting Room N602)

Abstract:

The combinatorial problem in collider experiments often makes it difficult to understand events and the corresponding underlying physics. The situation gets worse in the presence of the missing transverse momentum.  

In this talk, I will discuss various kinematic methods and machine learning approaches to address the combinatorial problem in SUSY- like events with two invisible particles at the LHC.  As a concrete example, I will focus on dileptonic tt ̄  events, where the combinatorial problem becomes an issue of binary classification: pairing the correct lepton with b quarks coming from the decays of the tops. 

Then I will consider the general case when the underlying mass spectrum is unknown, and hence no kinematic endpoint information is available. I will demonstrate that the efficiency for selecting the correct pairing is greatly improved by utilizing deep learning techniques, compared against existing methods based on kinematic variables.

Online Meeting room: https://meeting.tencent.com/dm/Xi4YKJuUzDvK

ID:  757-305-223   Password: 123456