ParticleNet: Jet Tagging via Particle Clouds

Abstract

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.

arXiv comment11 pages, 4 figures; v3: updated to match the version published in PRD; Code available at https://github.com/hqucms/ParticleNet

References in corpus (54)

• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• The anti-k_t jet clustering algorithm
• An Introduction to PYTHIA 8.2
• DELPHES 3, A modular framework for fast simulation of a generic collider experiment
• MXNet: A Flexible and Efficient Machine Learning Library for Heterogeneous Distributed Systems
• Particle-flow reconstruction and global event description with the CMS detector
• Identifying Boosted Objects with N-subjettiness
• A disciplined approach to neural network hyper-parameters: Part 1 -- learning rate, batch size, momentum, and weight decay
• Soft Drop
• Top-tagging: A Method for Identifying Boosted Hadronic Tops
• Energy Correlation Functions for Jet Substructure
• Jet reconstruction and performance using particle flow with the ATLAS Detector
• Maximizing Boosted Top Identification by Minimizing N-subjettiness
• Boosted objects: a probe of beyond the Standard Model physics
• Jet Substructure at the Large Hadron Collider: A Review of Recent Advances in Theory and Machine Learning
• Jet Substructure at the Tevatron and LHC: New results, new tools, new benchmarks
• Jet Substructure at the Large Hadron Collider: Experimental Review
• Classification without labels: Learning from mixed samples in high energy physics
• Deep learning in color: towards automated quark/gluon jet discrimination
• Jet-Images -- Deep Learning Edition
• Deep-learning Top Taggers or The End of QCD?
• Jet Substructure Classification in High-Energy Physics with Deep Neural Networks
• Energy Flow Networks: Deep Sets for Particle Jets
• The Machine Learning Landscape of Top Taggers
• Gaining (Mutual) Information about Quark/Gluon Discrimination
• Quark and Gluon Tagging at the LHC
• Boosted objects and jet substructure at the LHC
• Deep-learned Top Tagging with a Lorentz Layer
• Jet Flavor Classification in High-Energy Physics with Deep Neural Networks
• New Angles on Energy Correlation Functions
• Jet-Images: Computer Vision Inspired Techniques for Jet Tagging
• Systematics of quark/gluon tagging
• Quark and Gluon Jet Substructure
• Parton Shower Uncertainties in Jet Substructure Analyses with Deep Neural Networks
• Finding top quarks with shower deconstruction
• Pulling Out All the Tops with Computer Vision and Deep Learning
• The Lund Jet Plane
• QCD-Aware Recursive Neural Networks for Jet Physics
• JUNIPR: a Framework for Unsupervised Machine Learning in Particle Physics
• Learning to Classify from Impure Samples with High-Dimensional Data
• Casimir Meets Poisson: Improved Quark/Gluon Discrimination with Counting Observables
• Recursive Neural Networks in Quark/Gluon Tagging
• W-jet Tagging: Optimizing the Identification of Boosted Hadronically-Decaying W Bosons
• Jet Charge and Machine Learning
• Quark-Gluon Tagging: Machine Learning vs Detector
• How to Improve Top Tagging
• Quark-Gluon tagging with Shower Deconstruction: Unearthing dark matter and Higgs couplings
• An operational definition of quark and gluon jets
• Associated jet and subjet rates in light-quark and gluon jet discrimination
• Lorentz Boost Networks: Autonomous Physics-Inspired Feature Engineering
• Jet Constituents for Deep Neural Network Based Top Quark Tagging
• Infrared Safety of a Neural-Net Top Tagging Algorithm
• Long Short-Term Memory (LSTM) networks with jet constituents for boosted top tagging at the LHC
• Benchmarking an Even Better HEPTopTagger

Cited by in corpus (216)

• Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges
• Graph Neural Networks in Particle Physics
• Learning representations of irregular particle-detector geometry with distance-weighted graph networks
• An Efficient Lorentz Equivariant Graph Neural Network for Jet Tagging
• ABCNet: An attention-based method for particle tagging
• Anomaly detection with Convolutional Graph Neural Networks
• Neural Networks for Full Phase-space Reweighting and Parameter Tuning
• Graph Neural Networks for Particle Reconstruction in High Energy Physics detectors
• DisCo Fever: Robust Networks Through Distance Correlation
• MLPF: Efficient machine-learned particle-flow reconstruction using graph neural networks
• Modern Machine Learning and Particle Physics
• Interaction networks for the identification of boosted $H\to b\overline{b}$ decays
• Lorentz Group Equivariant Neural Network for Particle Physics
• A robust anomaly finder based on autoencoders
• FPGA-accelerated machine learning inference as a service for particle physics computing
• ABCDisCo: Automating the ABCD Method with Machine Learning
• Evaluating generative models in high energy physics
• A guide for deploying Deep Learning in LHC searches: How to achieve optimality and account for uncertainty
• Casting a graph net to catch dark showers
• Point Cloud Transformers applied to Collider Physics
• Mapping Machine-Learned Physics into a Human-Readable Space
• Supervised Jet Clustering with Graph Neural Networks for Lorentz Boosted Bosons
• Interpretable Deep Learning for Two-Prong Jet Classification with Jet Spectra
• Distance-Weighted Graph Neural Networks on FPGAs for Real-Time Particle Reconstruction in High Energy Physics
• PC-JeDi: Diffusion for Particle Cloud Generation in High Energy Physics
• Deep Set Auto Encoders for Anomaly Detection in Particle Physics
• Equivariant Energy Flow Networks for Jet Tagging
• Symmetries, Safety, and Self-Supervision
• Boosted $W/Z$ Tagging with Jet Charge and Deep Learning
• End-to-End Jet Classification of Quarks and Gluons with the CMS Open Data
• Per-Object Systematics using Deep-Learned Calibration
• Unsupervised Outlier Detection in Heavy-Ion Collisions
• Object condensation: one-stage grid-free multi-object reconstruction in physics detectors, graph and image data
• A fast centrality-meter for heavy-ion collisions at the CBM experiment
• The Boosted Higgs Jet Reconstruction via Graph Neural Network
• How to Understand Limitations of Generative Networks
• Unsupervised Hadronic SUEP at the LHC
• Graph Neural Networks for Charged Particle Tracking on FPGAs
• Measuring QCD Splittings with Invertible Networks
• Quarks and gluons in the Lund plane
• Towards a Computer Vision Particle Flow
• Search for Higgs boson decay to a charm quark-antiquark pair in proton-proton collisions at $\sqrt{s}$ = 13 TeV
• Energy-weighted Message Passing: an infra-red and collinear safe graph neural network algorithm
• Deep learning jet modifications in heavy-ion collisions
• Quantum Machine Learning for $b$-jet charge identification
• Towards Machine Learning Analytics for Jet Substructure
• Jet Flavour Tagging for Future Colliders with Fast Simulation
• Binary JUNIPR: an interpretable probabilistic model for discrimination
• Vector Boson Scattering Processes: Status and Prospects
• Learning the language of QCD jets with transformers
• Explainable Equivariant Neural Networks for Particle Physics: PELICAN
• On the Use of Neural Networks for Energy Reconstruction in High-granularity Calorimeters
• Lorentz group equivariant autoencoders
• Search for nonresonant pair production of highly energetic Higgs bosons decaying to bottom quarks
• Shared Data and Algorithms for Deep Learning in Fundamental Physics
• Machine Learning in High Energy Physics: A review of heavy-flavor jet tagging at the LHC
• Enriching the physics program of the CMS experiment via data scouting and data parking
• An equation-of-state-meter for CBM using PointNet
• Jet-Origin Identification and Its Application at an Electron-Positron Higgs Factory
• Non-Gaussianities in Collider Energy Flux
• Topological Reconstruction of Particle Physics Processes using Graph Neural Networks
• Unbinned Deep Learning Jet Substructure Measurement in High $Q^2$ ep collisions at HERA
• Boosting mono-jet searches with model-agnostic machine learning
• Secondary Vertex Finding in Jets with Neural Networks
• A Method to Simultaneously Facilitate All Jet Physics Tasks
• Machine Learning for Particle Flow Reconstruction at CMS
• Improved Constraints on Effective Top Quark Interactions using Edge Convolution Networks
• Search for a massive scalar resonance decaying to a light scalar and a Higgs boson in the four b quarks final state with boosted topology
• Reconstructing boosted Higgs jets from event image segmentation
• A Detailed Study of Interpretability of Deep Neural Network based Top Taggers
• Leveraging universality of jet taggers through transfer learning
• Searches for Higgs boson production through decays of heavy resonances
• Jet tagging algorithm of graph network with HaarPooling message passing
• Visualization for physics analysis improvement and applications in BESIII
• GPU coprocessors as a service for deep learning inference in high energy physics
• Resolving Extreme Jet Substructure
• A Lorentz-Equivariant Transformer for All of the LHC
• A Holistic Approach to Predicting Top Quark Kinematic Properties with the Covariant Particle Transformer
• Anomalies, Representations, and Self-Supervision
• Particle Convolution for High Energy Physics
• GNNBuilder: An Automated Framework for Generic Graph Neural Network Accelerator Generation, Simulation, and Optimization
• Deep Learning Jet Image as a Probe of Light Higgsino Dark Matter at the LHC
• Tagging more quark jet flavours at FCC-ee at 91 GeV with a transformer-based neural network
• Identification of Patterns in Cosmic-Ray Arrival Directions using Dynamic Graph Convolutional Neural Networks
• On Drell-Yan production of scalar leptoquarks coupling to heavy-quark flavours
• On the BSM reach of four top production at the LHC
• Graph Neural Networks for reconstruction and classification in KM3NeT
• Boosted top quark tagging and polarization measurement using machine learning
• Observation of a rare beta decay of the charmed baryon with a Graph Neural Network
• Flavor Physics at the CEPC: a General Perspective
• New Physics Search at the CEPC: a General Perspective
• New graph-neural-network flavor tagger for Belle II and measurement of $\sin 2{\varphi }_1$ in $B^0\to J/\psi K_{\text{S}}^0$ decays
• Jet Tagging with More-Interaction Particle Transformer
• Fast convolutional neural networks for identifying long-lived particles in a high-granularity calorimeter
• Maximum performance of strange-jet tagging at hadron colliders
• Graph Generative Models for Fast Detector Simulations in High Energy Physics
• Does Lorentz-symmetric design boost network performance in jet physics?
• Application of Graph Networks to background rejection in Imaging Air Cherenkov Telescopes
• Jet Discrimination with Quantum Complete Graph Neural Network
• A Tailored Convolutional Neural Network for Nonlinear Manifold Learning of Computational Physics Data using Unstructured Spatial Discretizations
• Real-time Graph Building on FPGAs for Machine Learning Trigger Applications in Particle Physics
• Learning Tree Structures from Leaves For Particle Decay Reconstruction
• Observation of the dead cone effect in charm and bottom quark jets and its QCD explanation
• Equivariant, Safe and Sensitive -- Graph Networks for New Physics
• Ultrafast jet classification on FPGAs for the HL-LHC
• Reconstruction of Unstable Heavy Particles Using Deep Symmetry-Preserving Attention Networks
• FAIR AI Models in High Energy Physics
• BIP: Boost Invariant Polynomials for Efficient Jet Tagging
• Semi-supervised Graph Neural Networks for Pileup Noise Removal
• Comparing Point Cloud Strategies for Collider Event Classification
• Is infrared-collinear safe information all you need for jet classification?
• Reconstruction of decays to merged photons using end-to-end deep learning with domain continuation in the CMS detector
• Search for Higgs boson pair production with one associated vector boson in proton-proton collisions at $\sqrt{s}$ = 13 TeV
• Learning Physics at Future $e^{-}{e}^{+}$ Colliders with Machine
• Hadronic Top Quark Polarimetry with ParticleNet
• Accuracy versus precision in boosted top tagging with the ATLAS detector
• Hypergraphs in LHC Phenomenology -- The Next Frontier of IRC-Safe Feature Extraction
• Using Machine Learning to Improve Neutron Identification in Water Cherenkov Detectors
• Deep Learning Jet Substructure from Two-Particle Correlation
• Search for vector-like leptons at a Muon Collider
• Aspen Open Jets: Unlocking LHC Data for Foundation Models in Particle Physics
• Search for dark QCD with emerging jets in proton-proton collisions at $\sqrt{s}$ = 13 TeV
• Jet Classification Using High-Level Features from Anatomy of Top Jets
• Particle Cloud Generation with Message Passing Generative Adversarial Networks
• Invariant Representation Driven Neural Classifier for Anti-QCD Jet Tagging
• Automatic detection of boosted Higgs boson and top quark jets in an event image
• Higgs and top physics reconstruction challenges and opportunities at FCC-ee
• Boosted top tagging and its interpretation using Shapley values
• Phenomenology at the Large Hadron Collider with Deep Learning: the case of vector-like quarks decaying to light jets
• Portable acceleration of CMS computing workflows with coprocessors as a service
• Interplay of Traditional Methods and Machine Learning Algorithms for Tagging Boosted Objects
• Performance of heavy-flavour jet identification in Lorentz-boosted topologies in proton-proton collisions at $\sqrt{s}$ = 13 TeV
• Unveiling the Secrets of New Physics Through Top Quark Tagging
• A rotation-equivariant graph neural network for learning hadronic SMEFT effects
• Rapidly Encoding Generalizable Dynamics in a Euclidean Symmetric Neural Network
• Attention to the strengths of physical interactions: Transformer and graph-based event classification for particle physics experiments
• Describing Hadronization via Histories and Observables for Monte-Carlo Event Reweighting
• Semi-Equivariant GNN Architectures for Jet Tagging
• Jet tagging in the Lund plane with graph networks
• Secondary Vertex Reconstruction with MaskFormers
• Domain-Adversarial Graph Neural Networks for $\Lambda$ Hyperon Identification with CLAS12
• One-to-one correspondence reconstruction at the electron-positron Higgs factory
• Classify the Higgs decays with the PFN and ParticleNet at electron-positron colliders
• Strange Jet Tagging
• Investigating top tagging with Y${}_{\text{m}}$-Splitter and N-subjettiness
• PAIReD jet: A multi-pronged resonance tagging strategy across all Lorentz boosts
• Cluster Counting Algorithm for the CEPC Drift Chamber using LSTM and DGCNN
• Foundations of automatic feature extraction at LHC--point clouds and graphs
• Study of WH production through vector boson scattering and extraction of the relative sign of the W and Z couplings to the Higgs boson in proton-proton collisions at $\sqrt{s}$ = 13 TeV
• Semi-supervised permutation invariant particle-level anomaly detection
• Fast multilabel classification of HEP constraints with deep learning
• Safety of Quark/Gluon Jet Classification
• Observation of the Singly Cabibbo-Suppressed Decay ${\Lambda }_c^{+}\to p{\pi }^0$
• Constraints on standard model effective field theory for a Higgs boson produced in association with W or Z bosons in the H $\to \mathrm{b}\bar{\mathrm{b}}$ decay channel in proton-proton collisions at $\sqrt{s}$ = 13 TeV
• Reconstruction of boosted and resolved multi-Higgs-boson events with symmetry-preserving attention networks
• Probing sub-TeV Higgsinos aided by a ML-based top tagger in the context of Trilinear RPV SUSY
• Fast Jet Tagging with MLP-Mixers on FPGAs
• Transforming jet flavour tagging at ATLAS
• Simultaneous probe of the charm and bottom quark Yukawa couplings using $\mathrm{t}\bar{\mathrm{t}}$H events
• Top-philic Machine Learning
• Jet energy calibration with deep learning as a Kubeflow pipeline
• LHC Hadronic Jet Generation Using Convolutional Variational Autoencoders with Normalizing Flows
• Tau lepton identification and reconstruction: a new frontier for jet-tagging ML algorithms
• BitHEP -- The Limits of Low-Precision ML in HEP
• Particle-level transformers for 95 GeV Higgs boson searches at future $e^{+}{e}^{-}$ Higgs factories
• Particle Graph Autoencoders and Differentiable, Learned Energy Mover's Distance
• Application of Graph Networks to a wide-field Water-Cherenkov-based Gamma-Ray Observatory
• Scalable Multi-Task Learning for Particle Collision Event Reconstruction with Heterogeneous Graph Neural Networks
• Search for dark matter production in association with a single top quark in proton-proton collisions at $\sqrt{s}$ = 13 TeV
• Guided Graph Compression for Quantum Graph Neural Networks
• Search for a heavy resonance decaying into a Z and a Higgs boson in events with an energetic jet and two electrons, two muons, or missing transverse momentum in proton-proton collisions at $\sqrt{s}$ = 13 TeV
• A method for correcting the substructure of multiprong jets using the Lund jet plane
• Identifying the Quantum Properties of Hadronic Resonances using Machine Learning
• JEDI-linear: Fast and Efficient Graph Neural Networks for Jet Tagging on FPGAs
• Tagging fully hadronic exotic decays of the vectorlike $\mathbf{B}$ quark using a graph neural network
• Particle Track Reconstruction using Geometric Deep Learning
• Interpretability Study on Deep Learning for Jet Physics at the Large Hadron Collider
• Image-Based Jet Analysis
• Performance studies of jet flavor tagging and measurement of $R_b(R_c)$ using ParticleNet at CEPC
• Search for resonances decaying to an anomalous jet and a Higgs boson in proton-proton collisions at $\sqrt{s}$ = 13 TeV
• OmniLearned: A Foundation Model Framework for All Tasks Involving Jet Physics
• Search for $\gamma$H production and constraints on the Yukawa couplings of light quarks to the Higgs boson
• Hadron Identification Prospects With Granular Calorimeters
• SuperSONIC: Cloud-Native Infrastructure for ML Inferencing
• Hybrid-graph neural network method for muon fast reconstruction in neutrino telescopes
• Choose Your Diffusion: Efficient and flexible ways to accelerate the diffusion model in fast high energy physics simulation
• Jet Reconstruction with Mamba Networks in Collider Events
• Search for $J/\psi \to K_S^0{K}_S^0$ and $\psi (3686)\to K_S^0{K}_S^0$
• Prospect for measurement of $CP$-violating observables in $B_s^0\to D_s^{\mp }{K}^{\pm }$ decays at a future $Z$ factory
• Reweighting and Analysing Event Generator Systematics by Neural Networks on High-Level Features
• Search for the radiative leptonic decay $D^{+}\to \gamma e^{+}{\nu }_e$ using Deep Learning
• A FAIR and AI-ready Higgs boson decay dataset
• Learnable cut flow for high energy physics
• Tagging ultra-boosted jets at FCC-hh using machine learning techniques
• Machine Learning Electroweakino Production
• Optimizing High Throughput Inference on Graph Neural Networks at Shared Computing Facilities with the NVIDIA Triton Inference Server
• Graph theory inspired anomaly detection at the LHC
• Vision Transformers for End-to-End Quark-Gluon Jet Classification from Calorimeter Images
• The Physics Behind ML-based Quark-Gluon Taggers
• Deep learning for jet modification in the presence of the QGP background
• Deep-learning jet flavor tagging for precision hadronic Higgs measurements at future $e^{+}{e}^{-}$ Higgs factories
• Probing new physics in the Boosted $HH\to b\bar{b}\gamma \gamma$ channel at the LHC
• Search for dark matter produced in association with a Higgs boson decaying to bottom quarks in proton-proton collisions at $\sqrt{s}$ = 13 TeV
• Combination of searches for heavy vector boson resonances in proton-proton collisions at $\sqrt{s}$ = 13 TeV
• jBOT: Semantic Jet Representation Clustering Emerges from Self-Distillation
• Theory, phenomenology, and experimental avenues for dark showers: a Snowmass 2021 report
• Search for pair production of heavy resonances in final states with a photon and large-radius jets in proton-proton collisions at $\sqrt{s}$ = 13 TeV
• Measurement of the jet mass in hadronic decays of boosted W bosons at 13 TeV and extraction of the W boson mass
• Search for Higgs boson production at high transverse momentum in the WW decay channel in proton-proton collisions at $\sqrt{s}$ = 13 TeV
• Observation of ${\Lambda }_c^{+}\to n{\pi }^{+}\eta$ and search for ${\Lambda }_c^{+}\to n{a}_0(980{)}^{+}$
• Search for Higgs boson pair production in the $\mathrm{b}\bar{\mathrm{b}}\mathrm{WW}$ decay channel with two leptons in the final state using proton-proton collision data at $\sqrt{s}$ = 13.6 TeV
• B-jet Tagging Using a Hybrid Edge Convolution and Transformer Architecture
• Machine learning tagged boosted dark photon: A signature of fermionic portal matter at the LHC
• Search for high-mass resonances in a final state comprising a gluon and two hadronically decaying W bosons in proton-proton collisions at $\sqrt{s}$ = 13 TeV
• Neutrino Reconstruction in TRIDENT Based on Graph Neural Network
• Explaining machine-learned particle-flow reconstruction