Boosting \(H \to b\bar{b}\) with machine learning
with Marat Freytsis, Ian Moult and Benjamin Nachman
Published J. High Energ. Phys. (2019) 2019: 181.
Journal link, Arxiv link

Developed Machine Learning methods (two-stream CNN model) to search for Higgs \(\to b\bar{b}\) quark decays, which is traditionally a difficult channel to search for because of the overwhelming QCD background from events such as \(gg \to b\bar{b}\). These searches also allow us to place tighter constraints on BSM limits with enough data, by probing the high-\(p_T\) regime.

Machine learning templates for QCD factorization in the search for physics beyond the standard model
with Wahid Bhimji and Benjamin Nachman
J. High Energ. Phys. (2019) 2019: 181.
Journal, Arxiv

Proposed an alternative to Jet Mass templates and other data-driven models for High Energy Physics searches, for instance for use in RPV-SUSY searches. Also proposed a modification to WGAN that greatly stabilises training and results for this problem, based off Lie theory.