News Item
Select label to Butter et al.
26 September 2018
"Deep-learned Top Tagging with a Lorentz Layer" by Anja Butter, Gregor Kasieczk, Tilman Plehn and Michael Russel has received the Select label.
Deep learning techniques based on low-level detector output are a promising new way to identify patterns in proton collision data at the LHC. They will, for example, allow us to identify the elementary particles making up so-called jets. This paper introduces a novel tagger that can identify boosted decaying top quarks using a set of measured four-momenta. It first shows that such taggers will outperform established techniques in a realistic detector environment. In addition, it traces the relevant patterns to the particle masses involved, extracted through the appropriate Minkowski metric.