Projects

This page is work in progress.

An overview of the research projects I’ve worked on at CMS, from my current PhD analysis back to my MSci project.

Higgs-tau CP analysis

Imperial College London, 2023–present

My PhD project is an analysis of the CP structure of the Yukawa coupling between the Higgs boson and tau leptons at the Compact Muon Solenoid (CMS) experiment, supported by the Schrödinger Scholarship. This work is ongoing — a fuller description, along with results and publications, will be added here as the analysis progresses.

Tau Identification (DeepTau)

Imperial College London / EPFL, November 2021–July 2022

My MSci project was on the identification of hadronic tau lepton decays at CMS, using a deep convolutional neural network called DeepTau. I introduced domain adaptation into the training workflow using adversarial machine learning techniques, which significantly reduced performance discrepancies between collider data and simulated events for discrimination against quark and gluon jets. This work has been integrated into central CMS software and used for physics analyses of the early Run 3 data-taking period.

My thesis is available online at cds.cern.ch/record/2827366/, you can read the abstract here:

“This thesis reports improved machine learning-based techniques to discriminate genuine decays of tau leptons into hadrons and a neutrino against all main backgrounds at the CMS experiment. The deep convolutional neural network, DeepTau, used for tau identification by physics analyses of the 2016-2018 data-taking period at CMS, shows a sizeably different performance on collider data versus Monte Carlo simulations. This effect is particularly prominent for regions of parameter space that have high genuine tau purity. The effects of this mismodelling on discrimination against quark and gluon jets are reduced by introducing domain adaptation into the training workflow. This approach was validated by comparing the performance of the resulting network on proton-proton collision data and simulated events. The use of these adversarial machine learning techniques reduced the discrepancies from 13.3% to 0.80% in the region where the purity of hadronic taus is expected to be above 96%, while having no significant impact on tau identification performance.”