Unbinned multivariate observables for global SMEFT analyses from machine learning

Abstract

Theoretical interpretations of particle physics data, such as the determination of the Wilson coefficients of the Standard Model Effective Field Theory (SMEFT), often involve the inference of multiple parameters from a global dataset. Optimizing such interpretations requires the identification of observables that exhibit the highest possible sensitivity to the underlying theory parameters. In this work we develop a flexible open source framework, ML4EFT, enabling the integration of unbinned multivariate observables into global SMEFT fits. As compared to traditional measurements, such observables enhance the sensitivity to the theory parameters by preventing the information loss incurred when binning in a subset of final-state kinematic variables. Our strategy combines machine learning regression and classification techniques to parameterize high-dimensional likelihood ratios, using the Monte Carlo replica method to estimate and propagate methodological uncertainties. As a proof of concept we construct unbinned multivariate observables for top-quark pair and Higgs+$Z$ production at the LHC, demonstrate their impact on the SMEFT parameter space as compared to binned measurements, and study the improved constraints associated to multivariate inputs. Since the number of neural networks to be trained scales quadratically with the number of parameters and can be fully parallelized, the ML4EFT framework is well-suited to construct unbinned multivariate observables which depend on up to tens of EFT coefficients, as required in global fits.

arXiv comment53 pages, 21 figures, the ML4EFT code is available from https://lhcfitnikhef.github.io/ML4EFT

References in corpus (65)

• The automated computation of tree-level and next-to-leading order differential cross sections, and their matching to parton shower simulations
• Asymptotic formulae for likelihood-based tests of new physics
• Parton distributions for the LHC Run II
• Dimension-Six Terms in the Standard Model Lagrangian
• Parton distributions with LHC data
• Automatized One-Loop Calculations in 4 and D dimensions
• Generating Feynman Diagrams and Amplitudes with FeynArts 3
• LHAPDF6: parton density access in the LHC precision era
• PDF4LHC recommendations for LHC Run II
• Importance Nested Sampling and the MultiNest Algorithm
• Renormalization Group Evolution of the Standard Model Dimension Six Operators III: Gauge Coupling Dependence and Phenomenology
• The Standard Model as an Effective Field Theory
• The Path to Proton Structure at One-Percent Accuracy
• A portrait of the Higgs boson by the CMS experiment ten years after the discovery
• A detailed map of Higgs boson interactions by the ATLAS experiment ten years after the discovery
• A posteriori inclusion of parton density functions in NLO QCD final-state calculations at hadron colliders: The APPLGRID Project
• The SMEFTsim package, theory and tools
• Top, Higgs, Diboson and Electroweak Fit to the Standard Model Effective Field Theory
• The Structure of the Proton in the LHC Precision Era
• A Monte Carlo global analysis of the Standard Model Effective Field Theory: the top quark sector
• Learning New Physics from a Machine
• O new physics, where art thou? A global search in the top sector
• Mining gold from implicit models to improve likelihood-free inference
• Constraining Effective Field Theories with Machine Learning
• Implications of Hadron Collider Observables on Parton Distribution Function Uncertainties
• A Guide to Constraining Effective Field Theories with Machine Learning
• The Gauge-Higgs Legacy of the LHC Run II
• SMEFTsim 3.0 -- a practical guide
• $\text{HEPfit}$: a Code for the Combination of Indirect and Direct Constraints on High Energy Physics Models
• Automation of the matrix element reweighting method
• Hadron structure in high-energy collisions
• aMCfast: automation of fast NLO computations for PDF fits
• Can New Physics hide inside the proton?
• The Matrix Element Method at Next-to-Leading Order
• Probing stop pair production at the LHC with graph neural networks
• Top and Beauty synergies in SMEFT-fits at present and future colliders
• Global and optimal probes for the top-quark effective field theory at future lepton colliders
• The Matrix Element Method and its Application to Measurements of the Top Quark Mass
• Higgs Physics: It ain't over till it's over
• New features in the JHU generator framework: constraining Higgs boson properties from on-shell and off-shell production
• Parton distributions in the SMEFT from high-energy Drell-Yan tails
• Unbiased determination of the proton structure function F_2^p with faithful uncertainty estimation
• SMEFT analysis of vector boson scattering and diboson data from the LHC Run II
• The Flavor of UV Physics
• NLO+NLL squark and gluino production cross-sections with threshold-improved parton distributions
• Top-pair production at the LHC with MiNNLO${}_{\mathrm{PS}}$
• PineAPPL: combining EW and QCD corrections for fast evaluation of LHC processes
• Extending the Matrix Element Method beyond the Born approximation: Calculating event weights at next-to-leading order accuracy
• Publishing statistical models: Getting the most out of particle physics experiments
• INFERNO: Inference-Aware Neural Optimisation
• Presenting Unbinned Differential Cross Section Results
• Optimized probes of $CP$-odd effects in the $t\bar{t}h$ process at hadron colliders
• Benchmarking simplified template cross sections in $WH$ production
• Learning from Radiation at a Very High Energy Lepton Collider
• Learning new physics efficiently with nonparametric methods
• Parametrized classifiers for optimal EFT sensitivity
• Parton Distributions and New Physics Searches: the Drell-Yan Forward-Backward Asymmetry as a Case Study
• The DNNLikelihood: enhancing likelihood distribution with Deep Learning
• Towards a new generation of parton densities with deep learning models
• Tree boosting for learning EFT parameters
• Optimal statistical inference in the presence of systematic uncertainties using neural network optimization based on binned Poisson likelihoods with nuisance parameters
• Next-to-next-to-leading order event generation for $VH$ production with $H\to b\bar{b}$ decay
• Machine learning of log-likelihood functions in global analysis of parton distributions
• A method for approximating optimal statistical significances with machine-learned likelihoods
• A data-based parametrization of parton distribution functions

Cited by in corpus (13)

• The top quark legacy of the LHC Run II for PDF and SMEFT analyses
• Simultaneous CTEQ-TEA extraction of PDFs and SMEFT parameters from jet and $t\bar{t}$ data
• Combining chromomagnetic and four-fermion operators with leading SMEFT operators for $gg\to hh$ at NLO QCD
• Constraining the Higgs Potential with Neural Simulation-based Inference for Di-Higgs Production
• Advancing Tools for Simulation-Based Inference
• Interpretable deep learning models for the inference and classification of LHC data
• A rotation-equivariant graph neural network for learning hadronic SMEFT effects
• Refinable modeling for unbinned SMEFT analyses
• Unbinned inclusive cross-section measurements with machine-learned systematic uncertainties
• 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
• Equivariant neural networks for robust $\text{CP}$ observables
• Designing Observables for Measurements with Deep Learning
• Data-Driven High-Dimensional Statistical Inference with Generative Models