ParticleNet: Jet Tagging via Particle Clouds
This addresses jet tagging in particle physics, providing a more efficient and symmetric representation for researchers in high-energy physics.
The paper tackled the problem of representing jets for machine learning in particle physics by proposing a 'particle cloud' approach, which achieved state-of-the-art performance on two jet tagging benchmarks with significant improvements over existing methods.
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.