Description
At the Belle II experiment, B meson decays can be studied with highest precision and in particular so called semi-leptonic decays where the B meson decays to a
hadron, a lepton and a neutrino. In this context, when the hadron contains an up quark (B → Xu l nu), important Standard Model parameters can be measured.
However, this process is overwhelmed by the much more likely decay to a hadron containing a charm quark (B → Xc l nu). Nowadays, most high energy physics
analyses make use of Machine Learning (ML) in order to improve the separation between signal and background. ML has already been used to distinguish B → Xu l
nu events from B → Xc l nu events at Belle II. Various algorithms can be compared in order to choose the most performant one. We propose the student to develop
a ML classifier (typically a Neural Network) for the B → Xu l nu analysis and compare its performance with other classifiers already used. Prior knowledge of
ML is not required (but could obviously help).
| Field | B1: Particle physics analysis (software-oriented) |
|---|---|
| DESY Place | Hamburg |
| DESY Division | FH |
| DESY Group | Belle II |
