SMuK 2023 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

T: Fachverband Teilchenphysik

T 29: Other Exp., EW

T 29.2: Vortrag

Dienstag, 21. März 2023, 17:15–17:30, HSZ/0103

A neural network for beam background decomposition in Belle II at SuperKEKB — •Yannik Buch, Ariane Frey, Lukas Herzberg, and Benjamin Schwenker — II. Physikalisches Institut, Georg-August-Universität Göttingen, Friedrich-Hund-Platz 1, 37077 Göttingen, Deutschland

The Belle II detector investigates the b-sector by measuring the decays of the Υ(4S) resonance. These Υ(4S) decays are produced by the SuperKEKB accelerator at KEK in Tsukuba, Japan. The goal of SuperKEKB is to achieve an instantaneous luminosity of 6.5×10³⁵cm⁻²s⁻¹, of which 4.7×10³⁴cm⁻²s⁻¹ has recently been reached.

The beam backgrounds at Belle II are mostly composed of storage and luminosity-induced backgrounds. Due to short beam lifetimes continuous top-up injections into both rings are necessary, resulting in injection-induced background spikes. BGNet is a neural network based diagnostic tool for real-time background decomposition and analysis. The training data for BGNet are 1Hz time series of diagnostic variables describing the state of the SuperKEKB collider subsystems. Using feature attribution to explain the predictions, provides clues to identify the most relevant causes of changes in background levels.