ParticleNet: Jet Tagging via Particle Clouds
By Huilin Qu, Loukas Gouskos. Physical Review D, 2019
324 citations recorded by the source corpus.
How to represent a jet is at the core of machine learning on jet physics. Inspired by the notion of point clouds, we propose a new approach that considers a jet as an unordered set of its constituent particles, effectively a ``particle cloud.'' Such a particle cloud representation of jets is efficient in incorporating raw information of jets and also explicitly respects the permutation symmetry. Based on the particle cloud representation, we propose ParticleNet, a customized neural network architecture using Dynamic Graph Convolutional Neural Network for jet tagging problems. The ParticleNet architecture achieves state-of-the-art performance on two representative jet tagging benchmarks and is improved significantly over existing methods.
Subfields: Artificial Intelligence, Nuclear and High Energy Physics
View Paper: https://www.semanticscholar.org/paper/58bcc1c9b60e761cea4d61fcecf4cff2f6216aa8
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Jonathan Shlomi
- Jean-Roch Vlimant
- I. Moult
- Benjamin Nachman
- Jiejun Xu
- Ross Maciejewski
- Peter W. Battaglia
- Si Zhang
In Nuclear and High Energy Physics
- I. Moult
- Benjamin Nachman
- A. Larkoski
- Jonathan Shlomi
- Peter W. Battaglia
- Jean-Roch Vlimant
- G. Soyez
- Jesse Thaler
Citations in the corpus
- Graph neural networks in particle physics (2020, methodology): “Their model outperforms the model from [52] on the quark/gluon classification task, indicating the importance of the attention mechanism — to which we come back to in sections 4.”
- Graph convolutional networks: a comprehensive review (2019, background): “ParticleNet, built upon edge convolutions [87], is a customized neural network architecture that operates directly on particle clouds for jet tagging [109].”
- Jet substructure at the Large Hadron Collider: A review of recent advances in theory and machine learning (2017, background): “Since that time, there have been advances in machine learning on point clouds (unordered variable-length sets) which have been applied to jet substructure [548, 549].”