Using neural networks for efficient evaluation of high multiplicity scattering amplitudes

Abstract

Precision theoretical predictions for high multiplicity scattering rely on the evaluation of increasingly complicated scattering amplitudes which come with an extremely high CPU cost. For state-of-the-art processes this can cause technical bottlenecks in the production of fully differential distributions. In this article we explore the possibility of using neural networks to approximate multi-variable scattering amplitudes and provide efficient inputs for Monte Carlo integration. We focus on QCD corrections to $e^{+}{e}^{-}\to$ jets up to one-loop and up to five jets. We demonstrate reliable interpolation when a series of networks are trained to amplitudes that have been divided into sectors defined by their infrared singularity structure. Complete simulations for one-loop distributions show speed improvements of at least an order of magnitude over a standard approach.

arXiv comment26 pages

References in corpus (27)

• Fusing Gauge Theory Tree Amplitudes Into Loop Amplitudes
• Three-jet cross sections to next-to-leading order
• Generalized Unitarity and One-Loop Amplitudes in N=4 Super-Yang-Mills
• Reducing full one-loop amplitudes to scalar integrals at the integrand level
• Automation of next-to-leading order computations in QCD: the FKS subtraction
• Full one-loop amplitudes from tree amplitudes
• Higgs boson pair production in gluon fusion at NLO with full top-quark mass dependence
• A Numerical Unitarity Formalism for Evaluating One-Loop Amplitudes
• Single-top production in MC@NLO
• Tops from Light Quarks: Full Mass Dependence at Two-Loops in QCD
• How to GAN LHC Events
• Numerical evaluation of virtual corrections to multi-jet production in massless QCD
• A proposal for a standard interface between Monte Carlo tools and one-loop programs
• Vegas Revisited: Adaptive Monte Carlo Integration Beyond Factorization
• Direct Extraction Of One Loop Rational Terms
• NLO QCD corrections to Higgs boson plus jet production with full top-quark mass dependence
• NLO predictions for Higgs boson pair production with full top quark mass dependence matched to parton showers
• Exploring phase space with Neural Importance Sampling
• Probing the trilinear Higgs boson coupling in di-Higgs production at NLO QCD including parton shower effects
• Machine Learning Uncertainties with Adversarial Neural Networks
• NLO QCD corrections to five-jet production at LEP and the extraction of alpha_s(M_Z)
• Foam: A General-Purpose Cellular Monte Carlo Event Generator
• Deep-Learning Jets with Uncertainties and More
• Lund jet images from generative and cycle-consistent adversarial networks
• A hierarchical phase space generator for QCD antenna structures
• SARGE: an algorithm for generating QCD-antennas
• (MC)**3 -- a Multi-Channel Markov Chain Monte Carlo algorithm for phase-space sampling

Cited by in corpus (35)

• Getting High: High Fidelity Simulation of High Granularity Calorimeters with High Speed
• Machine Learning and LHC Event Generation
• Invertible Networks or Partons to Detector and Back Again
• Understanding Event-Generation Networks via Uncertainties
• Accelerating Monte Carlo event generation -- rejection sampling using neural network event-weight estimates
• Event Generators for High-Energy Physics Experiments
• GANplifying Event Samples
• Generative Networks for Precision Enthusiasts
• A Neural Resampler for Monte Carlo Reweighting with Preserved Uncertainties
• A factorisation-aware Matrix element emulator
• Optimising simulations for diphoton production at hadron colliders using amplitude neural networks
• Targeting Multi-Loop Integrals with Neural Networks
• Phase Space Sampling and Inference from Weighted Events with Autoregressive Flows
• MadNIS -- Neural Multi-Channel Importance Sampling
• Les Houches 2021: Physics at TeV Colliders: Report on the Standard Model Precision Wishlist
• Per-Object Systematics using Deep-Learned Calibration
• Neural Network-based Top Tagger with Two-Point Energy Correlations and Geometry of Soft Emissions
• The MadNIS Reloaded
• One-loop matrix element emulation with factorisation awareness
• Comparing Machine Learning and Interpolation Methods for Loop-Level Calculations
• Differentiable MadNIS-Lite
• Hypergraphs in LHC Phenomenology -- The Next Frontier of IRC-Safe Feature Extraction
• Accurate Surrogate Amplitudes with Calibrated Uncertainties
• OASIS: Optimal Analysis-Specific Importance Sampling for event generation
• Learning Feynman integrals from differential equations with neural networks
• Learning the Simplicity of Scattering Amplitudes
• Accelerating multijet-merged event generation with neural network matrix element surrogates
• How to GAN Event Unweighting
• Predicting Feynman periods in ${\varphi }^4$-theory
• Extrapolating Jet Radiation with Autoregressive Transformers
• Amplitude Uncertainties Everywhere All at Once
• BitHEP -- The Limits of Low-Precision ML in HEP
• Optimising hadronic collider simulations using amplitude neural networks
• Amplitude Surrogates for Multi-Jet Processes
• A Novel Implementation of the Matrix Element Method at Next-to-Leading Order for the Measurement of the Higgs Self-Coupling ${\lambda }_{3H}$