MLPF: Efficient machine-learned particle-flow reconstruction using graph neural networks

Abstract

In general-purpose particle detectors, the particle-flow algorithm may be used to reconstruct a comprehensive particle-level view of the event by combining information from the calorimeters and the trackers, significantly improving the detector resolution for jets and the missing transverse momentum. In view of the planned high-luminosity upgrade of the CERN Large Hadron Collider (LHC), it is necessary to revisit existing reconstruction algorithms and ensure that both the physics and computational performance are sufficient in an environment with many simultaneous proton-proton interactions (pileup). Machine learning may offer a prospect for computationally efficient event reconstruction that is well-suited to heterogeneous computing platforms, while significantly improving the reconstruction quality over rule-based algorithms for granular detectors. We introduce MLPF, a novel, end-to-end trainable, machine-learned particle-flow algorithm based on parallelizable, computationally efficient, and scalable graph neural networks optimized using a multi-task objective on simulated events. We report the physics and computational performance of the MLPF algorithm on a Monte Carlo dataset of top quark-antiquark pairs produced in proton-proton collisions in conditions similar to those expected for the high-luminosity LHC. The MLPF algorithm improves the physics response with respect to a rule-based benchmark algorithm and demonstrates computationally scalable particle-flow reconstruction in a high-pileup environment.

arXiv comment15 pages, 10 figures

References in corpus (34)

• PYTHIA 6.4 Physics and Manual
• PYTHIA 6.2 Physics and Manual
• A Brief Introduction to PYTHIA 8.1
• DELPHES 3, A modular framework for fast simulation of a generic collider experiment
• Particle-flow reconstruction and global event description with the CMS detector
• Identification of heavy-flavour jets with the CMS detector in pp collisions at 13 TeV
• Fast inference of deep neural networks in FPGAs for particle physics
• ParticleNet: Jet Tagging via Particle Clouds
• Jet reconstruction and performance using particle flow with the ATLAS Detector
• ATLAS $b$-jet identification performance and efficiency measurement with $t\bar{t}$ events in $pp$ collisions at $\sqrt{s}=13$ TeV
• Energy Flow Networks: Deep Sets for Particle Jets
• Jet Flavour Classification Using DeepJet
• Graph Neural Networks in Particle Physics
• How to GAN LHC Events
• Identification of heavy, energetic, hadronically decaying particles using machine-learning techniques
• Learning representations of irregular particle-detector geometry with distance-weighted graph networks
• JEDI-net: a jet identification algorithm based on interaction networks
• The Metric Space of Collider Events
• Calorimetry with Deep Learning: Particle Simulation and Reconstruction for Collider Physics
• ABCNet: An attention-based method for particle tagging
• How to GAN away Detector Effects
• Interaction networks for the identification of boosted $H\to b\overline{b}$ decays
• The Tracking Machine Learning challenge : Accuracy phase
• FPGA-accelerated machine learning inference as a service for particle physics computing
• Measurement of the Diffractive Structure Function F_2^D(4) at HERA
• Supervised Jet Clustering with Graph Neural Networks for Lorentz Boosted Bosons
• Distance-Weighted Graph Neural Networks on FPGAs for Real-Time Particle Reconstruction in High Energy Physics
• Object condensation: one-stage grid-free multi-object reconstruction in physics detectors, graph and image data
• Identification of boosted Higgs bosons decaying into $b$-quark pairs with the ATLAS detector at 13 TeV
• HepSim: a repository with predictions for high-energy physics experiments
• Electromagnetic Showers Beyond Shower Shapes
• Measurement of sigma(ppbar -> Z + X) Br(Z -> tau+tau-) at sqrt(s)=1.96 TeV
• FPGAs-as-a-Service Toolkit (FaaST)
• GPU coprocessors as a service for deep learning inference in high energy physics

Cited by in corpus (38)

• Machine Learning and LHC Event Generation
• Applications and Techniques for Fast Machine Learning in Science
• Graph Neural Networks for Charged Particle Tracking on FPGAs
• Reconstructing particles in jets using set transformer and hypergraph prediction networks
• Shared Data and Algorithms for Deep Learning in Fundamental Physics
• The Tracking Machine Learning challenge : Throughput phase
• HHH Whitepaper
• Constraints on the trilinear and quartic Higgs couplings from triple Higgs production at the LHC and beyond
• Secondary Vertex Finding in Jets with Neural Networks
• Machine Learning for Particle Flow Reconstruction at CMS
• Improved Constraints on Effective Top Quark Interactions using Edge Convolution Networks
• GPU coprocessors as a service for deep learning inference in high energy physics
• A high-granularity calorimeter insert based on SiPM-on-tile technology at the future Electron-Ion Collider
• Optimizing Observables with Machine Learning for Better Unfolding
• Learning Uncertainties the Frequentist Way: Calibration and Correlation in High Energy Physics
• Does Lorentz-symmetric design boost network performance in jet physics?
• Learning Tree Structures from Leaves For Particle Decay Reconstruction
• Real-time Graph Building on FPGAs for Machine Learning Trigger Applications in Particle Physics
• Bias and Priors in Machine Learning Calibrations for High Energy Physics
• Semi-supervised Graph Neural Networks for Pileup Noise Removal
• Optimising longitudinal and lateral calorimeter granularity for software compensation in hadronic showers using deep neural networks
• Improved particle-flow event reconstruction with scalable neural networks for current and future particle detectors
• Learning to Isolate Muons
• CaloGraph: Graph-based diffusion model for fast shower generation in calorimeters with irregular geometry
• Interpretable deep learning models for the inference and classification of LHC data
• Higgs and top physics reconstruction challenges and opportunities at FCC-ee
• Portable acceleration of CMS computing workflows with coprocessors as a service
• Deep learning techniques for energy clustering in the CMS ECAL
• Set-Conditional Set Generation for Particle Physics
• Leveraging Staggered Tessellation for Enhanced Spatial Resolution in High-Granularity Calorimeters
• Parnassus: An Automated Approach to Accurate, Precise, and Fast Detector Simulation and Reconstruction
• Machine-learning-based prediction of parameters of secondaries in hadronic showers using calorimetric observables
• Fine-tuning machine-learned particle-flow reconstruction for new detector geometries in future colliders
• Foundations of automatic feature extraction at LHC--point clouds and graphs
• Hybrid Quantum Classical Graph Neural Networks for Particle Track Reconstruction
• Scalable Multi-Task Learning for Particle Collision Event Reconstruction with Heterogeneous Graph Neural Networks
• End-to-end multi-particle reconstruction in high occupancy imaging calorimeters with graph neural networks
• Distilling particle knowledge for fast reconstruction at high-energy physics experiments