Building on the notion of a particle physics detector as a camera and the collimated streams of high energy particles, or jets, it measures as an image, we investigate the potential of machine learning techniques based on deep learning architectures to identify highly boosted W bosons. Modern deep learning algorithms trained on jet images can out-perform standard physically-motivated feature driven approaches to jet tagging. We develop techniques for visualizing how these features are learned by the network and what additional information is used to improve performance. Finally, this interplay between physically-motivated feature driven tools and supervised learning algorithms is general and can be used to significantly increase the sensitivity to discover new particles and new forces, and gain a deeper understanding of the physics within jets.
de Oliveira, Luke, et al. "Jet-images — deep learning edition." Journal of High Energy Physics (Online), vol. 2016, no. 7, Jul. 2016. https://doi.org/10.1007/JHEP07(2016)069
de Oliveira, Luke, Kagan, Michael, Mackey, Lester, et al., "Jet-images — deep learning edition," Journal of High Energy Physics (Online) 2016, no. 7 (2016), https://doi.org/10.1007/JHEP07(2016)069
@article{osti_1271300,
author = {de Oliveira, Luke and Kagan, Michael and Mackey, Lester and Nachman, Benjamin and Schwartzman, Ariel},
title = {Jet-images — deep learning edition},
annote = {Building on the notion of a particle physics detector as a camera and the collimated streams of high energy particles, or jets, it measures as an image, we investigate the potential of machine learning techniques based on deep learning architectures to identify highly boosted W bosons. Modern deep learning algorithms trained on jet images can out-perform standard physically-motivated feature driven approaches to jet tagging. We develop techniques for visualizing how these features are learned by the network and what additional information is used to improve performance. Finally, this interplay between physically-motivated feature driven tools and supervised learning algorithms is general and can be used to significantly increase the sensitivity to discover new particles and new forces, and gain a deeper understanding of the physics within jets.},
doi = {10.1007/JHEP07(2016)069},
url = {https://www.osti.gov/biblio/1271300},
journal = {Journal of High Energy Physics (Online)},
issn = {ISSN 1029-8479},
number = {7},
volume = {2016},
place = {United States},
publisher = {Springer Berlin},
year = {2016},
month = {07}}
Kurth, Thorsten; Smorkalov, Mikhail; Deslippe, Jack
Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis on - SC '17https://doi.org/10.1145/3126908.3126916