Energy flow networks: deep sets for particle jets
A key question for machine learning approaches in particle physics is how to best represent and learn from collider events. As an event is intrinsically a variable-length unordered set of particles, we build upon recent machine learning efforts to learn directly from sets of features or “point cloud...
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
Springer Berlin Heidelberg
2019
|
| Online Access: | http://hdl.handle.net/1721.1/120141 |
| _version_ | 1811092940812976128 |
|---|---|
| author | Thaler, Jesse Komiske, Patrick T. Metodiev, Eric Mario |
| author2 | Massachusetts Institute of Technology. Center for Theoretical Physics |
| author_facet | Massachusetts Institute of Technology. Center for Theoretical Physics Thaler, Jesse Komiske, Patrick T. Metodiev, Eric Mario |
| author_sort | Thaler, Jesse |
| collection | MIT |
| description | A key question for machine learning approaches in particle physics is how to best represent and learn from collider events. As an event is intrinsically a variable-length unordered set of particles, we build upon recent machine learning efforts to learn directly from sets of features or “point clouds”. Adapting and specializing the “Deep Sets” framework to particle physics, we introduce Energy Flow Networks, which respect infrared and collinear safety by construction. We also develop Particle Flow Networks, which allow for general energy dependence and the inclusion of additional particle-level information such as charge and flavor. These networks feature a per-particle internal (latent) representation, and summing over all particles yields an overall event-level latent representation. We show how this latent space decomposition unifies existing event representations based on detector images and radiation moments. To demonstrate the power and simplicity of this set-based approach, we apply these networks to the collider task of discriminating quark jets from gluon jets, finding similar or improved performance compared to existing methods. We also show how the learned event representation can be directly visualized, providing insight into the inner workings of the model. These architectures lend themselves to efficiently processing and analyzing events for a wide variety of tasks at the Large Hadron Collider. Implementations and examples of our architectures are available online in our EnergyFlow package. Keywords:
Jets; QCD Phenomenology |
| first_indexed | 2024-09-23T15:36:31Z |
| format | Article |
| id | mit-1721.1/120141 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T15:36:31Z |
| publishDate | 2019 |
| publisher | Springer Berlin Heidelberg |
| record_format | dspace |
| spelling | mit-1721.1/1201412022-10-02T02:56:27Z Energy flow networks: deep sets for particle jets Thaler, Jesse Komiske, Patrick T. Metodiev, Eric Mario Massachusetts Institute of Technology. Center for Theoretical Physics Komiske, Patrick T. Metodiev, Eric Mario A key question for machine learning approaches in particle physics is how to best represent and learn from collider events. As an event is intrinsically a variable-length unordered set of particles, we build upon recent machine learning efforts to learn directly from sets of features or “point clouds”. Adapting and specializing the “Deep Sets” framework to particle physics, we introduce Energy Flow Networks, which respect infrared and collinear safety by construction. We also develop Particle Flow Networks, which allow for general energy dependence and the inclusion of additional particle-level information such as charge and flavor. These networks feature a per-particle internal (latent) representation, and summing over all particles yields an overall event-level latent representation. We show how this latent space decomposition unifies existing event representations based on detector images and radiation moments. To demonstrate the power and simplicity of this set-based approach, we apply these networks to the collider task of discriminating quark jets from gluon jets, finding similar or improved performance compared to existing methods. We also show how the learned event representation can be directly visualized, providing insight into the inner workings of the model. These architectures lend themselves to efficiently processing and analyzing events for a wide variety of tasks at the Large Hadron Collider. Implementations and examples of our architectures are available online in our EnergyFlow package. Keywords: Jets; QCD Phenomenology 2019-01-29T19:47:11Z 2019-01-29T19:47:11Z 2019-01 2019-01 2019-01-19T04:55:44Z Article http://purl.org/eprint/type/JournalArticle 1029-8479 http://hdl.handle.net/1721.1/120141 Komiske, Patrick T. et al. "Energy flow networks: deep sets for particle jets." Journal of High Energy Physics 2019 (January 2019): 121 © 2019 The Author(s) en https://doi.org/10.1007/JHEP01(2019)121 Journal of High Energy Physics Creative Commons Attribution https://creativecommons.org/licenses/by/4.0/ The Author(s) application/pdf Springer Berlin Heidelberg Springer Berlin Heidelberg |
| spellingShingle | Thaler, Jesse Komiske, Patrick T. Metodiev, Eric Mario Energy flow networks: deep sets for particle jets |
| title | Energy flow networks: deep sets for particle jets |
| title_full | Energy flow networks: deep sets for particle jets |
| title_fullStr | Energy flow networks: deep sets for particle jets |
| title_full_unstemmed | Energy flow networks: deep sets for particle jets |
| title_short | Energy flow networks: deep sets for particle jets |
| title_sort | energy flow networks deep sets for particle jets |
| url | http://hdl.handle.net/1721.1/120141 |
| work_keys_str_mv | AT thalerjesse energyflownetworksdeepsetsforparticlejets AT komiskepatrickt energyflownetworksdeepsetsforparticlejets AT metodievericmario energyflownetworksdeepsetsforparticlejets |