Weakly supervised classification in high energy physics

DOI10.1007/JHEP05(2017)145zbMATH Open1380.81439arXiv1702.00414OpenAlexW3101170768MaRDI QIDQ683055FDOQ683055

Authors: Lucio Mwinmaarong Dery, Benjamin Nachman, Francesco Rubbo, Ariel Schwartzman 

Publication date: 5 February 2018

Published in: Journal of High Energy Physics (Search for Journal in Brave)

Abstract: As machine learning algorithms become increasingly sophisticated to exploit subtle features of the data, they often become more dependent on simulations. This paper presents a new approach called weakly supervised classification in which class proportions are the only input into the machine learning algorithm. Using one of the most challenging binary classification tasks in high energy physics - quark versus gluon tagging - we show that weakly supervised classification can match the performance of fully supervised algorithms. Furthermore, by design, the new algorithm is insensitive to any mis-modeling of discriminating features in the data by the simulation. Weakly supervised classification is a general procedure that can be applied to a wide variety of learning problems to boost performance and robustness when detailed simulations are not reliable or not available.

Full work available at URL: https://arxiv.org/abs/1702.00414

Recommendations

• AN EMPIRICAL STUDY OF BOOSTED NEURAL NETWORK FOR PARTICLE CLASSIFICATION IN HIGH ENERGY COLLISIONS
• Supervised Deep Learning in High Energy Phenomenology: a Mini Review*
• Weakly supervised clustering: learning fine-grained signals from coarse labels
• Nonparametric semi-supervised classification with application to signal detection in high energy physics
• Automated feature selection procedure for particle jet classification

zbMATH Keywords

jetsmachine learning algorithmweakly supervised classification

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05) Strong interaction, including quantum chromodynamics (81V05)

Cites Work

• FastJet user manual (for version 3.0.2)
• A brief introduction to PYTHIA 8.1
• PYTHIA 6.4 physics and manual
• The anti-\(k_t\) jet clustering algorithm
• A general framework for implementing NLO calculations in shower Monte Carlo programs: the POWHEG BOX
• The automated computation of tree-level and next-to-leading order differential cross sections, and their matching to parton shower simulations
• Solving the multiple instance problem with axis-parallel rectangles.
• Title not available (Why is that?)
• Multiple instance classification: review, taxonomy and comparative study
• Deep learning in color: towards automated quark/gluon jet discrimination

Cited In (7)

• Supervised Deep Learning in High Energy Phenomenology: a Mini Review*
• INFERNO: inference-aware neural optimisation
• On asymptotic charges in 3D gravity
• Nonparametric semi-supervised classification with application to signal detection in high energy physics
• Learning to inflate. A gradient ascent approach to random inflation
• Automated feature selection procedure for particle jet classification
• AN EMPIRICAL STUDY OF BOOSTED NEURAL NETWORK FOR PARTICLE CLASSIFICATION IN HIGH ENERGY COLLISIONS

Uses Software

• POWHEG BOX
• PYTHIA8
• SHERPA
• FastJet
• Herwig++
• Keras
• Adam
• GitHub

This page was built for publication: Weakly supervised classification in high energy physics

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q683055)