Uncertainties associated with GAN-generated datasets in high energy physics - INSPIRE

DataBETA

Uncertainties associated with GAN-generated datasets in high energy physics

Feb 14, 2020

25 pages

Published in:

SciPost Phys. 12 (2022) 3, 104

Published: Mar 24, 2022

e-Print:

2002.06307 [hep-ph]

DOI:

10.21468/SciPostPhys.12.3.104

Report number:

FERMILAB-PUB-21-663-QIS

View in:

reference search35 citations

Citations per year

Abstract: (SciPost Fundation)

Recently, Generative Adversarial Networks (GANs) trained on samples of traditionally simulated collider events have been proposed as a way of generating larger simulated datasets at a reduced computational cost. In this paper we point out that data generated by a GAN cannot statistically be better than the data it was trained on, and critically examine the applicability of GANs in various situations, including a) for replacing the entire Monte Carlo pipeline or parts of it, and b) to produce datasets for usage in highly sensitive analyses or sub-optimal ones. We present our arguments using information theoretic demonstrations, a toy example, as well as in the form of a formal statement, and identify some potential valid uses of GANs in collider simulations.

Note:

28 pages, 11 figures. Revised submission to SciPost

network
costs
energy: high
Monte Carlo
statistical analysis

References(65)

Figures(9)

[1]

Planning the Future of U.S. Particle Physics (Snowmass 2013): Chapter 9: Computing Frontier

et al.

e-Print:
- 1401.6117

[2]

Fermilab Computing at the Intensity Frontier

et al.

- J.Phys.Conf.Ser. 664 (2015) 3, 032012
•
DOI:
- 10.1088/1742-6596/664/3/032012

[3]

A Roadmap for HEP Software and Computing R&D for the 2020s

HEP Software Foundation

Collaboration

•

Johannes Albrecht
(
- Dortmund U.
)

et al.

- Comput.Softw.Big Sci. 3 (2019) 1, 7
•
e-Print:
- 1712.06982
•
DOI:
- 10.1007/s41781-018-0018-8

[4]

Computational challenges for MC event generation

Andy Buckley
(
- Glasgow U.
)

- J.Phys.Conf.Ser. 1525 (2020) 1, 012023
•
e-Print:
- 1908.00167
•
DOI:
- 10.1088/1742-6596/1525/1/012023

[5]

New strategies of the LHC experiments to meet the computing requirements of the HL-LHC era

Dagmar Adamova
(
- Rez, Nucl. Phys. Inst.
)
,
Maarten Litmaath
(
- CERN
)

- PoS BORMIO2017 (2017) 053
•
DOI:
- 10.22323/1.302.0053

[6]

CMS Computing Resources: Meeting the Demands of the High-Luminosity LHC Physics Program

David Lange
(
- Princeton U.
)
,
Kenneth Bloom
(
- Nebraska U.
)
,
Tommaso Boccali
(
- INFN, Pisa and
- Pisa U.
)
,
Oliver Gutsche
(
- Fermilab
)
,
Eric Vaandering
(
- Fermilab
)

- EPJ Web Conf. 214 (2019) 03055
•
DOI:
- 10.1051/epjconf/201921403055

[7]

Generative Adversarial Networks

et al.

e-Print:
- 1406.2661

[8]

From Lagrangians to Events: Computer Tutorial at the MC4BSM-2012 Workshop

Stefan Ask
(
- Cambridge U.
)
,
Neil D. Christensen
(
- Pittsburgh U.
)
,
Claude Duhr
(
- Zurich, ETH
)
,
Christophe Grojean
(
- CERN
)
,
Stefan Hoeche
(
- SLAC
)

et al.

e-Print:
- 1209.0297

[9]

Constraining the Parameters of High-Dimensional Models with Active Learning

Sascha Caron
(
- Nijmegen U., IMAPP and
- Nikhef, Amsterdam
)
,
Tom Heskes
(
- Nijmegen U.
)
,
Sydney Otten
(
- Nijmegen U., IMAPP and
- U. Amsterdam, GRAPPA
)
,
Bob Stienen
(
- Nijmegen U., IMAPP
)

- Eur.Phys.J.C 79 (2019) 11, 944
•
e-Print:
- 1905.08628
•
DOI:
- 10.1140/epjc/s10052-019-7437-5

[10]

Fast Data-Driven Simulation of Cherenkov Detectors Using Generative Adversarial Networks

LHCb

Collaboration

•

Artem Maevskiy
(
- Higher Sch. of Economics, Moscow
)

et al.

- J.Phys.Conf.Ser. 1525 (2020) 1, 012097
•
e-Print:
- 1905.11825
•
DOI:
- 10.1088/1742-6596/1525/1/012097

[11]

Calorimetry with deep learning: particle simulation and reconstruction for collider physics

Dawit Belayneh
(
- U. Chicago (main)
)
,
Federico Carminati
(
- CERN
)
,
Amir Farbin
(
- Texas U., Arlington (main)
)
,
Benjamin Hooberman
(
- Illinois U., Urbana (main)
)
,
Gulrukh Khattak
(
- CERN and
- Unlisted, PK
)

et al.

- Eur.Phys.J.C 80 (2020) 7, 688
•
e-Print:
- 1912.06794
•
DOI:
- 10.1140/epjc/s10052-020-8251-9

[12]

Large-Scale Distributed Training Applied to Generative Adversarial Networks for Calorimeter Simulation

Jean-Roch Vlimant
(
- Caltech
)
,
Felice Pantaleo
(
- CERN
)
,
Maurizio Pierini
(
- CERN
)
,
Vladimir Loncar
(
- CERN
)
,
Sofia Vallecorsa
(
- CERN and
- Gangneung-Wonju Natl. U.
)

et al.

- EPJ Web Conf. 214 (2019) 06025
•
DOI:
- 10.1051/epjconf/201921406025

[13]

Simulating the LHCb hadron calorimeter with generative adversarial networks

- Nuovo Cim.C 42 (2019) 4, 197
•
DOI:
- 10.1393/ncc/i2019-19197-3

[14]

Learning Particle Physics by Example: Location-Aware Generative Adversarial Networks for Physics Synthesis

Luke de Oliveira
(
- LBNL, Berkeley
)
,
Michela Paganini
(
- LBNL, Berkeley and
- Yale U.
)
,
Benjamin Nachman
(
- LBNL, Berkeley
)

- Comput.Softw.Big Sci. 1 (2017) 1, 4
•
e-Print:
- 1701.05927
•
DOI:
- 10.1007/s41781-017-0004-6

[15]

Lund jet images from generative and cycle-consistent adversarial networks

Stefano Carrazza
(
- INFN, Milan and
- Milan U.
)
,
Frédéric A. Dreyer
(
- Oxford U.
)

- Eur.Phys.J.C 79 (2019) 11, 979
•
e-Print:
- 1909.01359
•
DOI:
- 10.1140/epjc/s10052-019-7501-1

[16]

Accelerating Science with Generative Adversarial Networks: An Application to 3D Particle Showers in Multilayer Calorimeters

Michela Paganini
(
- LBNL, Berkeley and
- Yale U.
)
,
Luke de Oliveira
(
- LBNL, Berkeley
)
,
Benjamin Nachman
(
- LBNL, Berkeley
)

- Phys.Rev.Lett. 120 (2018) 4, 042003
•
e-Print:
- 1705.02355
•
DOI:
- 10.1103/PhysRevLett.120.042003

[17]

Controlling Physical Attributes in GAN-Accelerated Simulation of Electromagnetic Calorimeters

- J.Phys.Conf.Ser. 1085 (2018) 4, 042017
•
e-Print:
- 1711.08813
•
DOI:
- 10.1088/1742-6596/1085/4/042017

[18]

CaloGAN : Simulating 3D high energy particle showers in multilayer electromagnetic calorimeters with generative adversarial networks

Michela Paganini
(
- Yale U. and
- LBL, Berkeley
)
,
Luke de Oliveira
(
- LBL, Berkeley
)
,
Benjamin Nachman
(
- LBL, Berkeley
)

- Phys.Rev.D 97 (2018) 1, 014021
•
e-Print:
- 1712.10321
•
DOI:
- 10.1103/PhysRevD.97.014021

[19]

Three dimensional Generative Adversarial Networks for fast simulation

et al.

- J.Phys.Conf.Ser. 1085 (2018) 3, 032016
•
DOI:
- 10.1088/1742-6596/1085/3/032016

[20]

Generating and refining particle detector simulations using the Wasserstein distance in adversarial networks

Martin Erdmann
(
- Aachen, Tech. Hochsch.
)
,
Lukas Geiger
(
- Aachen, Tech. Hochsch.
)
,
Jonas Glombitza
(
- Aachen, Tech. Hochsch.
)
,
David Schmidt
(
)

- Comput.Softw.Big Sci. 2 (2018) 1, 4
•
e-Print:
- 1802.03325
•
DOI:
- 10.1007/s41781-018-0008-x

[21]

Fast and Accurate Simulation of Particle Detectors Using Generative Adversarial Networks

Pasquale Musella
(
- ETH, Zurich (main)
)
,
Francesco Pandolfi
(
- INFN, Rome
)

- Comput.Softw.Big Sci. 2 (2018) 1, 8
•
e-Print:
- 1805.00850
•
DOI:
- 10.1007/s41781-018-0015-y

[22]

Precise simulation of electromagnetic calorimeter showers using a Wasserstein Generative Adversarial Network

Martin Erdmann
(
- Aachen, Tech. Hochsch.
)
,
Jonas Glombitza
(
- Aachen, Tech. Hochsch.
)
,
Thorben Quast
(
- Aachen, Tech. Hochsch. and
- CERN
)

- Comput.Softw.Big Sci. 3 (2019) 1, 4
•
e-Print:
- 1807.01954
•
DOI:
- 10.1007/s41781-018-0019-7

[23]

3D convolutional GAN for fast simulation

Sofia Vallecorsa
(
- CERN and
- Gangneung-Wonju Natl. U.
)
,
Federico Carminati
(
- CERN
)
,
Gulrukh Khattak
(
- CERN and
- Peshawar U.
)

- EPJ Web Conf. 214 (2019) 02010
•
DOI:
- 10.1051/epjconf/201921402010

[24]

Event Generation and Statistical Sampling for Physics with Deep Generative Models and a Density Information Buffer

Sydney Otten
(
- Nijmegen U., IMAPP and
- U. Amsterdam, GRAPPA
)
,
Sascha Caron
(
- Nijmegen U., IMAPP and
- Nikhef, Amsterdam
)
,
Wieske de Swart
(
- Nijmegen U., IMAPP
)
,
Melissa van Beekveld
(
- Nijmegen U., IMAPP
)
,
Luc Hendriks
(
- Nijmegen U., IMAPP
)

et al.

- Nature Commun. 12 (2021) 1, 2985
•
e-Print:
- 1901.00875
•
DOI:
- 10.1038/s41467-021-22616-z

[25]

How to GAN LHC Events

- SciPost Phys. 7 (2019) 6, 075
•
e-Print:
- 1907.03764
•
DOI:
- 10.21468/SciPostPhys.7.6.075

1-25 of 65
1
2
3
25 / page