Machine Learning for High-Energy Collider Physics
Fundamental physics, in particular high-energy collider physics, seeks to understand the natural world at the smallest scales, leading experimentally to the creation of large, complex datasets. Machine learning comprises a powerful set of statistical and computational tools enabling comprehensive ex...
Main Author: | |
---|---|
Other Authors: | |
Format: | Thesis |
Published: |
Massachusetts Institute of Technology
2022
|
Online Access: | https://hdl.handle.net/1721.1/142702 |
_version_ | 1811086908092055552 |
---|---|
author | Komiske III, Patrick Theodore |
author2 | Thaler, Jesse |
author_facet | Thaler, Jesse Komiske III, Patrick Theodore |
author_sort | Komiske III, Patrick Theodore |
collection | MIT |
description | Fundamental physics, in particular high-energy collider physics, seeks to understand the natural world at the smallest scales, leading experimentally to the creation of large, complex datasets. Machine learning comprises a powerful set of statistical and computational tools enabling comprehensive exploitation of data. In this thesis, I develop machine learning methods to facilitate cutting-edge analysis techniques in particle physics. I model collider events as point clouds and develop neural network architectures that respect the inherent permutation symmetry and variable number of particles of an event, with infrared safety naturally incorporated. I further design a procedure that uses high-dimensional classifiers to achieve full-phase space, unbinned unfolding of all observables simultaneously. In the second part of this thesis, I define a distance metric between collider events based on optimal transport that allows for a rigorous construction of "event space" and its corresponding geometry. Using public datasets provided by the CMS collaboration, I explore this metric on a dataset of real jets, demonstrating its viability as an experimental method as well as the value of public collider data in benchmarking new techniques. |
first_indexed | 2024-09-23T13:36:46Z |
format | Thesis |
id | mit-1721.1/142702 |
institution | Massachusetts Institute of Technology |
last_indexed | 2024-09-23T13:36:46Z |
publishDate | 2022 |
publisher | Massachusetts Institute of Technology |
record_format | dspace |
spelling | mit-1721.1/1427022022-05-25T03:19:48Z Machine Learning for High-Energy Collider Physics Komiske III, Patrick Theodore Thaler, Jesse Massachusetts Institute of Technology. Department of Physics Fundamental physics, in particular high-energy collider physics, seeks to understand the natural world at the smallest scales, leading experimentally to the creation of large, complex datasets. Machine learning comprises a powerful set of statistical and computational tools enabling comprehensive exploitation of data. In this thesis, I develop machine learning methods to facilitate cutting-edge analysis techniques in particle physics. I model collider events as point clouds and develop neural network architectures that respect the inherent permutation symmetry and variable number of particles of an event, with infrared safety naturally incorporated. I further design a procedure that uses high-dimensional classifiers to achieve full-phase space, unbinned unfolding of all observables simultaneously. In the second part of this thesis, I define a distance metric between collider events based on optimal transport that allows for a rigorous construction of "event space" and its corresponding geometry. Using public datasets provided by the CMS collaboration, I explore this metric on a dataset of real jets, demonstrating its viability as an experimental method as well as the value of public collider data in benchmarking new techniques. Ph.D. 2022-05-24T19:20:06Z 2022-05-24T19:20:06Z 2021-06 2022-05-19T23:48:26.851Z Thesis https://hdl.handle.net/1721.1/142702 0000-0002-2983-9518 In Copyright - Educational Use Permitted Copyright MIT http://rightsstatements.org/page/InC-EDU/1.0/ application/pdf Massachusetts Institute of Technology |
spellingShingle | Komiske III, Patrick Theodore Machine Learning for High-Energy Collider Physics |
title | Machine Learning for High-Energy Collider Physics |
title_full | Machine Learning for High-Energy Collider Physics |
title_fullStr | Machine Learning for High-Energy Collider Physics |
title_full_unstemmed | Machine Learning for High-Energy Collider Physics |
title_short | Machine Learning for High-Energy Collider Physics |
title_sort | machine learning for high energy collider physics |
url | https://hdl.handle.net/1721.1/142702 |
work_keys_str_mv | AT komiskeiiipatricktheodore machinelearningforhighenergycolliderphysics |