Jul 16 – 21, 2017
Embassy Suites Buffalo
US/Eastern timezone

Identification of boosted top quarks and W bosons with Machine learning in ATLAS

Jul 19, 2017, 3:00 PM
Embassy Suites Buffalo

Embassy Suites Buffalo

200 Delaware Avenue Buffalo, NY 14202


Oliver Majersky (Comenius University (SK))


We present techniques for the identification of hadronically-decaying W bosons and top quarks using high-level features as inputs to boosted decision trees and deep neural networks in the ATLAS experiment at sqrt(s)=13 TeV. The performance of these machine learning based taggers is compared in Monte Carlo simulation with various different tagging algorithms. An improvement in background rejection with respect to different taggers is observed. In addition, the performance of the machine learning taggers is examined in full Run-II data set in top quark pair, dijet and photon+jet topologies.

Presentation materials