Khurana, Ishan;
(2024)
Mitigating Accidental Coincidence Backgrounds in the LZ experiment: demonstration of a Machine Learning Approach using first data.
Doctoral thesis (Ph.D), UCL (University College London).
Preview |
Text
Khurana_10185176_thesis.pdf Download (26MB) | Preview |
Abstract
Understanding the nature of dark matter has been one of the pre-eminent problems in particle physics, given the compelling body of astrophysical and cosmological evidence pointing towards the existence and abundance of an elusive, invisible form of matter. A theoretically well motivated candidate is the Weakly Interacting Massive Particle (WIMP). LUX-ZEPLIN (LZ) is a direct detection experiment searching for WIMP interactions with xenon nuclei. Based in the Sanford Underground Research Facility (SURF) in Lead, South Dakota, it monitors a 7 tonne active volume of liquid xenon and began its first science run (SR1) on the 23rd of December 2021 consisting of a 60 live day exposure using a fiducial mass of 5.5 tonnes. First results from the SR1 data report a world-leading limit for spin-independent scattering at 36 GeV/c 2 , rejecting cross sections above σSI = 9.2×10−48 cm2 at the 90% confidence level. With increased access to computing power, another area of research that has grown significantly in the last decade is that of applied machine learning (ML). Realising the potential gains in physics analyses that can be made using ML techniques requires a meshing of expertise in both detector physics and ML tools. The research reported in this thesis presents a framework for background rejection in the LZ data using Boosted Decision Trees (BDT). The effectiveness of the approach is demonstrated on accidental coincidence events, though it can be generalised to other analyses or ML models. Crucially, the BDT based rejection of accidental coincidence events reduces the rate of such backgrounds by a factor of four compared with the official SR1 analysis, at a relatively small loss in signal efficiency. Thus it brings the accidentals rate to a level required for LZ to reach its projected sensitivity.
| Type: | Thesis (Doctoral) |
|---|---|
| Qualification: | Ph.D |
| Title: | Mitigating Accidental Coincidence Backgrounds in the LZ experiment: demonstration of a Machine Learning Approach using first data |
| Open access status: | An open access version is available from UCL Discovery |
| Language: | English |
| Additional information: | Copyright © The Author 2024. Original content in this thesis is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) Licence (https://creativecommons.org/licenses/by-nc/4.0/). Any third-party copyright material present remains the property of its respective owner(s) and is licensed under its existing terms. Access may initially be restricted at the author’s request. |
| UCL classification: | UCL UCL > Provost and Vice Provost Offices > UCL BEAMS UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Maths and Physical Sciences |
| URI: | https://discovery-pp.ucl.ac.uk/id/eprint/10185176 |
Archive Staff Only
 |
View Item |
