Modern Machine Learning and Particle Physics

@article{Schwartz2021ModernML,
  title={Modern Machine Learning and Particle Physics},
  author={Matthew D. Schwartz},
  journal={Harvard Data Science Review},
  year={2021}
}
  • M. Schwartz
  • Published 1 March 2021
  • Physics
  • Harvard Data Science Review
Over the past five years, modern machine learning has been quietly revoltionizing particle physics. Old methodology is being outdated and entirely new ways of thinking about data are becoming commonplace. This article will review some aspects of the natural synergy between modern machine learning and particle physics, focusing on applications at the Large Hadron Collider. A sampling of examples is given, from signal/background discrimination tasks using supervised learning to direct data-driven… 

Figures from this paper

Uncertainty-aware machine learning for high energy physics

TLDR
This work studies the use of a classifier that is fully aware of uncertainties and their corresponding nuisance parameters and shows that this dependence can actually enhance the sensitivity to parameters of interest.

Boosted decision trees in the era of new physics: a smuon analysis case study

Abstract Machine learning algorithms are growing increasingly popular in particle physics analyses, where they are used for their ability to solve difficult classification and regression problems.

Metalearning and data augmentation for mass-generalized jet taggers

TLDR
A simple data augmentation strategy that standardises the angular scale of jets with different masses is sufficient to produce strong generalisation, and the meta-learning algorithms provide only a small improvement in generalisation when combined with this augmentation.

A machine learning global analysis approach to enhance the precision of the Higgs decay branching fractions in electron-positron colliders

Several high energy e+e− colliders are proposed as Higgs factories by the international high energy physics community. One of the most important goals of these projects is to study the Higgs

Energy reconstruction for large liquid scintillator detectors with machine learning techniques: aggregated features approach

TLDR
This work presents machine learning methods for energy reconstruction in JUNO, the most advanced detector of its type, and focuses on positron events in the energy range of 0–10 MeV which corresponds to the main signal inJUNO — neutrinos originated from nuclear reactor cores and detected via an inverse beta-decay channel.

Graph Neural Networks in Particle Physics: Implementations, Innovations, and Challenges

TLDR
A range of graph neural networks capabilities that allow a wide variety of high- and low-level physical features to be attached to measurements and, by the same token, to best match unique GNN capabilities to unique HEP obstacles are presented.

A global analysis approach to improve the precision of the Higgs decay branching fractions at electron-positron colliders

Several high energy e+e− colliders are proposed as Higgs factories by the international high energy physics community. One of the most important goals of these projects is to study the Higgs

On scientific understanding with artificial intelligence

Imagine an oracle that correctly predicts the outcome of every particle physics experiment, the products of every chemical reaction, or the function of every protein. Such an oracle would

Translating neutron star observations to nuclear symmetry energy via artificial neural networks

One of the most significant challenges involved in efforts to understand the equation of state of dense neutron-rich matter is the uncertain density dependence of the nuclear symmetry energy. In

Translating Neutron Star Observations to Nuclear Symmetry Energy via Deep Neural Networks

One of the most significant challenges involved in efforts to understand the equation of state of dense neutron-rich matter is the uncertain density dependence of the nuclear symmetry energy. In

References

SHOWING 1-10 OF 25 REFERENCES

The Machine Learning landscape of top taggers

TLDR
Based on the established task of identifying boosted, hadronically decaying top quarks, a wide range of modern machine learning approaches are compared, finding that these new approaches are extremely powerful and great fun.

Automating the ABCD method with machine learning

TLDR
Automating the ABCD method with machine learning significantly improves performance in terms of ABCD closure, background rejection and signal contamination, and uncovers a previously unappreciated aspect of the method: its accuracy hinges on having low signal contamination in control regions not just overall, but relative to the signal fraction in the signal region.

Classification without labels: learning from mixed samples in high energy physics

TLDR
This paper introduces the paradigm of classification without labels (CWoLa) in which a classifier is trained to distinguish statistical mixtures of classes, which are common in collider physics.

Data science applications to string theory

Learning to classify from impure samples with high-dimensional data

TLDR
This work demonstrates that complex, high-dimensional classifiers can also be trained on impure mixtures using weak supervision techniques, with performance comparable to what could be achieved with pure samples.

Jet Tagging via Particle Clouds

TLDR
This work proposes ParticleNet, a customized neural network architecture using Dynamic Graph Convolutional Neural Network for jet tagging problems and achieves state-of-the-art performance on two representative jet tagging benchmarks and is improved significantly over existing methods.

Deep-learning top taggers or the end of QCD?

A bstractMachine learning based on convolutional neural networks can be used to study jet images from the LHC. Top tagging in fat jets offers a well-defined framework to establish our DeepTop

A Convolutional Neural Network Neutrino Event Classifier

TLDR
This algorithm, CVN (Convolutional Visual Network) identifies neutrino interactions based on their topology without the need for detailed reconstruction and outperforms algorithms currently in use by the NOvA collaboration.

UvA-DARE (Digital Academic Repository) Performance of top-quark and W-boson tagging with ATLAS in Run 2 of the LHC

The performance of identification algorithms (“taggers”) for hadronically decaying top quarks and W bosons in pp collisions at √ s = 13 TeV recorded by the ATLAS experiment at the Large Hadron

binary junipr: An Interpretable Probabilistic Model for Discrimination.

TLDR
This Letter shows how the training of the separate junipr models can be refined in the context of classification to optimize discrimination power and achieves state-of-the-art performance for quark-gluon discrimination and top tagging.