How to GAN away detector effects

Bellagente, Marco; Butter, Anja; Kasieczka, G.; Plehn, Tilman; Winterhalder, Ramon

doi:10.21468/scipostphys.8.4.070

Cited by 80 publications

(87 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this case the (unweighted) events are 4-momenta of external particles. We ignore all information on the particle identification, except for its mass, which allows us to reduce external 4-momenta to external 3momenta [15,30]. Because the input events might have been object to detector effects we do not assume energy-momentum conservation for the entire event.…”

Section: Lhc Eventsmentioning

confidence: 99%

“…We require a minimal p T of 10 GeV, a maximal rapidity of 2.5 for each electron, and a minimal angular separation of 0.4. We do not apply a detector simulation at this stage, because our focus is on comparing the generated and true distributions, and we have shown that detector simulations can be included trivially in our GAN setup [15,30].…”

Section: Background Subtractionmentioning

confidence: 99%

“…This includes for instance phase space integration [11], event generation [12][13][14][15], detector simulations [16][17][18][19][20][21][22], unfolding [23], parton showers [24][25][26][27][28], or searches for physics beyond the Standard Model [29]. Most recently, we have shown that fully conditional GANs can be used to invert typical Monte Carlo processes at the LHC, like for instance a fast detector simulation [30].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

How to GAN event subtraction

2020

Self Cite

View full text Add to dashboard Cite

Subtracting event samples is a common task in LHC simulation and analysis, and standard solutions tend to be inefficient. We employ generative adversarial networks to produce new event samples with a phase space distribution corresponding to added or subtracted input samples. We first illustrate for a toy example how such a network beats the statistical limitations of the training data. We then show how such a network can be used to subtract background events or to include non-local collinear subtraction events at the level of unweighted 4-vector events.

show abstract

Section: Lhc Eventsmentioning

confidence: 99%

Section: Background Subtractionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

How to GAN event subtraction

2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…(4.22), modern unfolding techniques are necessary; see, for example, refs. [119][120][121][122][123] and references therein. Such techniques "correct" reconstructed distributions/observables for real, detector-level and analysis-level acceptance efficiencies, enabling more direct comparisons to truth-level, MC predictions [119,120].…”

Section: Impact Of Selection Cuts On Polarized Distributions As Notementioning

confidence: 99%

Automated predictions from polarized matrix elements

Franzosi

Mattelaer

Ruiz

et al. 2020

J. High Energ. Phys.

View full text Add to dashboard Cite

The anticipated experimental resolution and data cache of the High Luminosity Large Hadron Collider will enable precision investigations of polarization in multiboson processes. This includes, for the first time, vector boson scattering. To facilitate such studies, we report the automation of polarized matrix element computations in the publicly available Monte Carlo tool suite, MadGraph5 aMC@NLO. This enables scattering and decay simulations involving helicity-polarized asymptotic or intermediate states, preserving both spin-correlation and off-shell effects. As demonstrations of the method, we investigate the leading order production and decay of polarized weak gauge bosons in the process pp → jjW + λ W − λ , with helicity eigenstates (λ, λ) defined in various reference frames. We consider the Standard Model at both O(α 4) and O(α 2 α 2 s) as well as a benchmark composite Higgs scenario. We report good agreement with polarization studies based on the On-Shell Projection (OSP) technique. Future capabilities are discussed.

show abstract

“…Interest in deep learning in collider physics [1][2][3][4][5] has been growing in recent years. Many applications of deep learning have appeared in jet classification , anomaly detection [27][28][29][30][31][32][33][34][35][36][37], particle identification [38][39][40], pileup mitigation [41][42][43], event generation [44][45][46][47][48][49][50][51][52][53][54][55][56][57][58], unfolding [59,60], and parton distribution functions [61][62][63][64][65][66][67][68][69][70][71]…”

Section: Introductionmentioning

confidence: 99%

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

Chakraborty

Lim

Nojiri

et al. 2020

J. High Energ. Phys.

View full text Add to dashboard Cite

Deep neural networks trained on jet images have been successful in classifying different kinds of jets. In this paper, we identify the crucial physics features that could reproduce the classification performance of the convolutional neural network in the top jet vs. QCD jet classification. We design a neural network that considers two types of substructural features: two-point energy correlations, and the IRC unsafe counting variables of a morphological analysis of jet images. The new set of IRC unsafe variables can be described by Minkowski functionals from integral geometry. To integrate these features into a single framework, we reintroduce two-point energy correlations in terms of a graph neural network and provide the other features to the network afterward. The network shows a comparable classification performance to the convolutional neural network. Since both networks are using IRC unsafe features at some level, the results based on simulations are often dependent on the event generator choice. We compare the classification results of Pythia 8 and Herwig 7, and a simple reweighting on the distribution of IRC unsafe features reduces the difference between the results from the two simulations.

show abstract

How to GAN away detector effects

Cited by 80 publications

References 46 publications

How to GAN event subtraction

How to GAN event subtraction

Automated predictions from polarized matrix elements

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

Contact Info

Product

Resources

About