Combine and Conquer: Event Reconstruction with Bayesian Ensemble Neural Networks

Abstract

Ensemble learning is a technique where multiple component learners are combined through a protocol. We propose an Ensemble Neural Network (ENN) that uses the combined latent-feature space of multiple neural network classifiers to improve the representation of the network hypothesis. We apply this approach to construct an ENN from Convolutional and Recurrent Neural Networks to discriminate top-quark jets from QCD jets. Such ENN provides the flexibility to improve the classification beyond simple prediction combining methods by linking different sources of error correlations, hence improving the representation between data and hypothesis. In combination with Bayesian techniques, we show that it can reduce epistemic uncertainties and the entropy of the hypothesis by simultaneously exploiting various kinematic correlations of the system, which also makes the network less susceptible to a limitation in training sample size.

arXiv comment23 pages, 17 figures. Accepted version for publication in JHEP

References in corpus (47)

• The anti-k_t jet clustering algorithm
• An Introduction to PYTHIA 8.2
• FastJet user manual
• DELPHES 3, A modular framework for fast simulation of a generic collider experiment
• Jet substructure as a new Higgs search channel at the LHC
• Jet Substructure at the Large Hadron Collider: A Review of Recent Advances in Theory and Machine Learning
• Fat Jets for a Light Higgs
• Stop Reconstruction with Tagged Tops
• 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
• Learning New Physics from a Machine
• Deep-learned Top Tagging with a Lorentz Layer
• Constraining Effective Field Theories with Machine Learning
• Quark and Gluon Jet Substructure
• Adversarially-trained autoencoders for robust unsupervised new physics searches
• Finding top quarks with shower deconstruction
• Finding physics signals with shower deconstruction
• Pulling Out All the Tops with Computer Vision and Deep Learning
• QCD-Aware Recursive Neural Networks for Jet Physics
• Enhanced Higgs to ${\tau }^{+}{\tau }^{-}$ Searches with Deep Learning
• Boosting $H\to b\bar{b}$ with Machine Learning
• Recursive Neural Networks in Quark/Gluon Tagging
• Machine Learning Uncertainties with Adversarial Neural Networks
• Finding physics signals with event deconstruction
• A Neural Resampler for Monte Carlo Reweighting with Preserved Uncertainties
• How to Improve Top Tagging
• Deep-Learning Jets with Uncertainties and More
• A guide for deploying Deep Learning in LHC searches: How to achieve optimality and account for uncertainty
• Mapping Machine-Learned Physics into a Human-Readable Space
• Supervised Jet Clustering with Graph Neural Networks for Lorentz Boosted Bosons
• Automating the Construction of Jet Observables with Machine Learning
• Infrared Safety of a Neural-Net Top Tagging Algorithm
• The Cambridge Jet algorithm: features and applications
• Convolutional Neural Networks for Direct Detection of Dark Matter
• Reports of My Demise Are Greatly Exaggerated: $N$-subjettiness Taggers Take On Jet Images
• Stacking machine learning classifiers to identify Higgs bosons at the LHC
• Spectral Analysis of Jet Substructure with Neural Networks: Boosted Higgs Case
• Investigating the Topology Dependence of Quark and Gluon Jets
• Machine Learning Templates for QCD Factorization in the Search for Physics Beyond the Standard Model
• Enhancing searches for resonances with machine learning and moment decomposition
• Towards recognizing the light facet of the Higgs Boson
• Applying Rule Ensembles to the Search for Super-Symmetry at the Large Hadron Collider
• Sensing Higgs cascade decays through memory

Cited by in corpus (18)

• Autoencoders for unsupervised anomaly detection in high energy physics
• Understanding Event-Generation Networks via Uncertainties
• Uncertainty Aware Learning for High Energy Physics
• The forgotten channels: charged Higgs boson decays to a $W^{\pm }$ and a non-SM-like Higgs boson
• Energy-weighted Message Passing: an infra-red and collinear safe graph neural network algorithm
• Classical versus Quantum: comparing Tensor Network-based Quantum Circuits on LHC data
• Quantum-probabilistic Hamiltonian learning for generative modelling & anomaly detection
• Improved Constraints on Effective Top Quark Interactions using Edge Convolution Networks
• Active learning BSM parameter spaces
• Learning Uncertainties the Frequentist Way: Calibration and Correlation in High Energy Physics
• Searches for new physics with boosted top quarks in the MadAnalysis 5 and Rivet frameworks
• Quantum-inspired event reconstruction with Tensor Networks: Matrix Product States
• Bias and Priors in Machine Learning Calibrations for High Energy Physics
• Is infrared-collinear safe information all you need for jet classification?
• Calibrating Bayesian Generative Machine Learning for Bayesiamplification
• Interpretable deep learning models for the inference and classification of LHC data
• Foundations of automatic feature extraction at LHC--point clouds and graphs
• Graph theory inspired anomaly detection at the LHC