SMuK 2023 – wissenschaftliches Programm

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

T: Fachverband Teilchenphysik

T 15: Neutrinos, Dark Matter III

T 15.3: Vortrag

Montag, 20. März 2023, 17:00–17:15, POT/0006

Low-frequency noise classification for the SuperCDMS experiment using Machine Learning — •Sukeerthi Dharani for the SuperCDMS collaboration — Karlsruhe Institute of Technology, Institute for Astroparticle Physics — University of Hamburg, Institute for Experimental Physics

The SuperCDMS Soudan experiment was a direct dark matter search experiment that was operated from 2012 to 2015 at the Soudan Underground Laboratory in Minnesota, USA. It used germanium crystal detectors at cryogenic temperatures to search for dark matter-nucleon scattering events. The experiment was affected by broadband low-frequency (LF) noise due to vibrations from the cryocooler, which deteriorated the detector baseline resolution and increased the noise trigger rate. The LF noise events can have a similar pulse shape as the low-energy signal events, making it difficult to remove them at low energies. In the final low ionization threshold analysis, this has led to stronger event selection criteria to remove LF noise events which set a higher analysis threshold and thus reduced the sensitivity of the experiment to low-mass dark matter. Currently, an LF noise selection criterion using machine learning is being studied. Under investigation is a convolutional neural network that yields better signal purity while also retaining signal efficiency. This talk discusses the machine learning-based classification of LF noise and its preliminary results.