Matrix Element Regression with Deep Neural Networks -- breaking the CPU barrier

Abstract

The Matrix Element Method (MEM) is a powerful method to extract information from measured events at collider experiments. Compared to multivariate techniques built on large sets of experimental data, the MEM does not rely on an examples-based learning phase but directly exploits our knowledge of the physics processes. This comes at a price, both in term of complexity and computing time since the required multi-dimensional integral of a rapidly varying function needs to be evaluated for every event and physics process considered. This can be mitigated by optimizing the integration, as is done in the MoMEMta package, but the computing time remains a concern, and often makes the use of the MEM in full-scale analysis unpractical or impossible. We investigate in this paper the use of a Deep Neural Network (DNN) built by regression of the MEM integral as an ansatz for analysis, especially in the search for new physics.

arXiv comment20 pages, 18 figures

References in corpus (39)

• Adam: A Method for Stochastic Optimization
• The automated computation of tree-level and next-to-leading order differential cross sections, and their matching to parton shower simulations
• Deep Learning using Rectified Linear Units (ReLU)
• L2 Regularization for Learning Kernels
• Parameterized Machine Learning for High-Energy Physics
• Search for the Standard Model Higgs boson produced in association with top quarks and decaying into $b\bar{b}$ in pp collisions at $\sqrt{s}$ = 8 TeV with the ATLAS detector
• Mining gold from implicit models to improve likelihood-free inference
• Evidence for the associated production of the Higgs boson and a top quark pair with the ATLAS detector
• Approximating Likelihood Ratios with Calibrated Discriminative Classifiers
• Comparison of non-linear activation functions for deep neural networks on MNIST classification task
• JUNIPR: a Framework for Unsupervised Machine Learning in Particle Physics
• Measurement of the Higgs boson coupling properties in the $H\to Z{Z}^{\ast }\to 4\ell$ decay channel at $\sqrt{s}$ = 13 TeV with the ATLAS detector
• i-flow: High-dimensional Integration and Sampling with Normalizing Flows
• Search for the Standard Model Higgs boson produced in association with top quarks and decaying into a $b\bar{b}$ pair in $pp$ collisions at $\sqrt{s}$ = 13 TeV with the ATLAS detector
• Exploring phase space with Neural Importance Sampling
• Weakly Supervised Classification in High Energy Physics
• Learning to Pivot with Adversarial Networks
• Search for a heavy Higgs boson decaying into a $Z$ boson and another heavy Higgs boson in the $\ell \ell bb$ final state in $pp$ collisions at $\sqrt{s}=13$ TeV with the ATLAS detector
• Evidence for associated production of a Higgs boson with a top quark pair in final states with electrons, muons, and hadronically decaying $\tau$ leptons at $\sqrt{s}=$ 13 TeV
• Search for $\mathrm{t}\overline{\mathrm{t}}$H production in the H $\to$ $\mathrm{b}\overline{\mathrm{b}}$ decay channel with leptonic $\mathrm{t}\overline{\mathrm{t}}$ decays in proton-proton collisions at $\sqrt{s}=$ 13 TeV
• Evidence for single top-quark production in the $s$-channel in proton-proton collisions at $\sqrt{s}=$8 TeV with the ATLAS detector using the Matrix Element Method
• Neural Network-Based Approach to Phase Space Integration
• Measurement of the top quark mass with dilepton events selected using neuroevolution at CDF
• Search for new neutral Higgs bosons through the H $\to$ ZA $\to$ ${\ell }^{+}{\ell }^{-}\mathrm{b}\bar{\mathrm{b}}$ process in pp collisions at $\sqrt{s}=$ 13 TeV
• Search for $\mathrm{t}\overline{\mathrm{t}}$H production in the all-jet final state in proton-proton collisions at $\sqrt{s}=$ 13 TeV
• Measurement of spin correlations in t t-bar production using the matrix element method in the muon + jets final state in pp collisions at sqrt(s) = 8 TeV
• Efficient Monte Carlo Integration Using Boosted Decision Trees and Generative Deep Neural Networks
• Semi-Supervised Anomaly Detection - Towards Model-Independent Searches of New Physics
• LHC analysis-specific datasets with Generative Adversarial Networks
• Test of Analysis Method for Top-Antitop Production and Decay Events
• The DNNLikelihood: enhancing likelihood distribution with Deep Learning
• Accelerated Matrix Element Method with Parallel Computing
• Precision measurement of the top-quark mass in lepton+jets final states
• Likelihood-free inference with an improved cross-entropy estimator
• Measurements of the Top-quark Mass and the ttbar Cross Section in the Hadronic tau + Jets Decay Channel at sqrt(s) = 1.96 TeV
• Measurement of the Top Quark Mass Using the Matrix Element Technique in Dilepton Final States
• Top Quark Mass Measurement in the tt-bar All Hadronic Channel using a Matrix Element Technique in ppbar Collisions at sqrt s = 1.96 TeV
• MoMEMta, a modular toolkit for the Matrix Element Method at the LHC
• Measurement of the top-quark mass in the lepton+jets channel using a matrix element technique with the CDF II detector

Cited by in corpus (8)

• Accelerating Monte Carlo event generation -- rejection sampling using neural network event-weight estimates
• A factorisation-aware Matrix element emulator
• Two Invertible Networks for the Matrix Element Method
• Lorentz group equivariant autoencoders
• Returning CP-Observables to The Frames They Belong
• Precision-Machine Learning for the Matrix Element Method
• Interpretable deep learning models for the inference and classification of LHC data
• A Novel Implementation of the Matrix Element Method at Next-to-Leading Order for the Measurement of the Higgs Self-Coupling ${\lambda }_{3H}$