Reference : QCD-Aware Recursive Neural Networks for Jet Physics
Scientific journals : Article
Physical, chemical, mathematical & earth Sciences : Physics
Engineering, computing & technology : Computer science
QCD-Aware Recursive Neural Networks for Jet Physics
Louppe, Gilles mailto [Université de Liège - ULiège > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Big Data >]
Cho, Kyunghyun [> >]
Becot, Cyril [> >]
Cranmer, Kyle [> >]
Journal of High Energy Physics
Yes (verified by ORBi)
[en] High Energy Physics - Phenomenology ; Physics - Data Analysis ; Statistics and Probability ; Statistics - Machine Learning
[en] Recent progress in applying machine learning for jet physics has been built upon an analogy between calorimeters and images. In this work, we present a novel class of recursive neural networks built instead upon an analogy between QCD and natural languages. In the analogy, four-momenta are like words and the clustering history of sequential recombination jet algorithms is like the parsing of a sentence. Our approach works directly with the four-momenta of a variable-length set of particles, and the jet-based tree structure varies on an event-by-event basis. Our experiments highlight the flexibility of our method for building task-specific jet embeddings and show that recursive architectures are significantly more accurate and data efficient than previous image-based networks. We extend the analogy from individual jets (sentences) to full events (paragraphs), and show for the first time an event-level classifier operating on all the stable particles produced in an LHC event.
11 pages, 5 figures, corresponding code at

File(s) associated to this reference

Fulltext file(s):

Open access
1702.00748.pdfAuthor preprint536.02 kBView/Open

Bookmark and Share SFX Query

All documents in ORBi are protected by a user license.