Neural network-based top tagger with two-point energy correlations and geometry of soft emissions

Chakraborty, Amit; Lim, S.; Nojiri, Mihoko M.; Takeuchi, Michihisa

doi:10.1007/jhep07(2020)111

Cited by 32 publications

(23 citation statements)

References 146 publications

(212 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We refer to refs. [62][63][64][65][66][67][68] for examples of deep learning architectures that incorporate specific physics features to guide event classification.…”

Section: Discussionmentioning

confidence: 99%

Casting a graph net to catch dark showers

et al. 2021

View full text Add to dashboard Cite

Strongly interacting dark sectors predict novel LHC signatures such as semi-visible jets resulting from dark showers that contain both stable and unstable dark mesons. Distinguishing such semi-visible jets from large QCD backgrounds is difficult and constitutes an exciting challenge for jet classification. In this article we explore the potential of supervised deep neural networks to identify semi-visible jets. We show that dynamic graph convolutional neural networks operating on so-called particle clouds outperform convolutional neural networks analysing jet images as well as other neural networks based on Lorentz vectors. We investigate how the performance depends on the properties of the dark shower and discuss training on mixed samples as a strategy to reduce model dependence. By modifying an existing mono-jet analysis we show that LHC sensitivity to dark sectors can be enhanced by more than an order of magnitude by using the dynamic graph network as a dark shower tagger.

show abstract

“…We refer to refs. [62][63][64][65][66][67][68] for examples of deep learning architectures that incorporate specific physics features to guide event classification.…”

Section: Discussionmentioning

confidence: 99%

Casting a graph net to catch dark showers

et al. 2021

View full text Add to dashboard Cite

show abstract

“…This topic has been the focus of much attention over the past decade, with a range of approaches being developed to extract information from a jet's substructure [2][3][4][5][6]. In recent years, a new generation of tools based on deep learning models have emerged, which can achieve very high performance on specific benchmarks [7][8][9][10][11][12][13][14][15][16][17][18][19][20] and provide some insights into what kinematic variables drive the discrimination performance [21][22][23][24][25][26][27][28][29][30]. A limitation of such deep learning-based methods is the difficulty to estimate their uncertainties, as well as their proneness to rely on unphysical features present in the training data to achieve their high performance, as this data is generally derived from Monte Carlo simulations of proton collisions.…”

Section: Introductionmentioning

confidence: 99%

Jet tagging in the Lund plane with graph networks

Dreyer¹,

2021

J. High Energ. Phys.

View full text Add to dashboard Cite

The identification of boosted heavy particles such as top quarks or vector bosons is one of the key problems arising in experimental studies at the Large Hadron Collider. In this article, we introduce LundNet, a novel jet tagging method which relies on graph neural networks and an efficient description of the radiation patterns within a jet to optimally disentangle signatures of boosted objects from background events. We apply this framework to a number of different benchmarks, showing significantly improved performance for top tagging compared to existing state-of-the-art algorithms. We study the robustness of the LundNet taggers to non-perturbative and detector effects, and show how kinematic cuts in the Lund plane can mitigate overfitting of the neural network to model-dependent contributions. Finally, we consider the computational complexity of this method and its scaling as a function of kinematic Lund plane cuts, showing an order of magnitude improvement in speed over previous graph-based taggers.

show abstract

“…The substructure is created by the interspersing of visible hadrons with dark hadrons. The substructure observables which are least affected by model dependence can be used in searches, and also as inputs to machine learning algorithms trying to identify semi-visible jet via anomaly detection [34][35][36][37][38][39][40][41], assuming the relatively similar contribution from signal and background processes.…”

Section: Discussionmentioning

confidence: 99%

Exploring jet substructure in semi-visible jets

Kar

Sinha

2021

SciPost Phys.

View full text Add to dashboard Cite

Semi-visible jets arise in strongly interacting dark sectors, where parton evolution includes dark sector emissions, resulting in jets overlapping with missing transverse momentum. The implementation of semi-visible jets is done using the Pythia Hidden valley module to duplicate the QCD sector showering. In this work, several jet substructure observables have been examined to compare semi-visible jets (signal) and light quark/gluon jets (background). These comparisons were performed using different dark hadron fractions in the semi-visible jets. The extreme scenarios where signal consists either of entirely dark hadrons or visible hadrons offers a chance to understand the effect of the specific dark shower model employed in these comparisons. We attempt to decouple the behaviour of jet-substructure observables due to inherent semi-visible jet properties, from model dependence owing to the existence of only one dark shower model as mentioned above.

show abstract

Neural network-based top tagger with two-point energy correlations and geometry of soft emissions

Cited by 32 publications

References 146 publications

Casting a graph net to catch dark showers

Casting a graph net to catch dark showers

Jet tagging in the Lund plane with graph networks

Exploring jet substructure in semi-visible jets

Contact Info

Product

Resources

About