國立虎尾科技大學 |

Domain generalization with machine learning in the NOvA experiment

Record Type:	Language materials, printed : Monograph/item
Title/Author:	Domain generalization with machine learning in the NOvA experiment/ by Andrew T.C. Sutton.
Author:	Sutton, Andrew T. C.
Published:	Cham :Springer Nature Switzerland : : 2023.,
Description:	xi, 170 p. :illustrations (chiefly color), digital ; : 24 cm.;
Notes:	"Doctoral Thesis accepted by the University of Virginia, USA."
Contained By:	Springer Nature eBook
Subject:	Neutrinos. -
Online resource:	https://doi.org/10.1007/978-3-031-43583-6
ISBN:	9783031435836

Domain generalization with machine learning in the NOvA experiment
Sutton, Andrew T. C.

Domain generalization with machine learning in the NOvA experiment[electronic resource] /by Andrew T.C. Sutton. - Cham :Springer Nature Switzerland :2023. - xi, 170 p. :illustrations (chiefly color), digital ;24 cm. - Springer theses,2190-5061. - Springer theses..

"Doctoral Thesis accepted by the University of Virginia, USA."

Chapter 1: Neutrinos: A Desperate Remedy -- Chapter 2. A Review of Neutrino Physics -- Chapter 3. The NOvA Experiment -- Chapter 4. Event Reconstruction -- Chapter 5. The 3-Flavor Analysis -- Chapter 6. A Long Short-Term Memory Neural Network -- Chapter 7. Domain Generalization by Adversarial Training -- Chapter 8. Conclusion.

This thesis presents significant advances in the use of neural networks to study the properties of neutrinos. Machine learning tools like neural networks (NN) can be used to identify the particle types or determine their energies in detectors such as those used in the NOvA neutrino experiment, which studies changes in a beam of neutrinos as it propagates approximately 800 km through the earth. NOvA relies heavily on simulations of the physics processes and the detector response; these simulations work well, but do not match the real experiment perfectly. Thus, neural networks trained on simulated datasets must include systematic uncertainties that account for possible imperfections in the simulation. This thesis presents the first application in HEP of adversarial domain generalization to a regression neural network. Applying domain generalization to problems with large systematic variations will reduce the impact of uncertainties while avoiding the risk of falsely constraining the phase space. Reducing the impact of systematic uncertainties makes NOvA analysis more robust, and improves the significance of experimental results.

ISBN: 9783031435836

Standard No.: 10.1007/978-3-031-43583-6doiSubjects--Topical Terms:

786066
Neutrinos.

LC Class. No.: QC793.5.N42

Dewey Class. No.: 539.7215

Domain generalization with machine learning in the NOvA experiment
LDR:02612nam a2200349 a 4500 001 1119955
003 DE-He213
005 20231108154304.0
006 m d
007 cr nn 008maaau
008 240612s2023 sz s 0 eng d
020 $a 9783031435836 $q (electronic bk.)
020 $a 9783031435829 $q (paper)
024 7 $a 10.1007/978-3-031-43583-6 $2 doi
035 $a 978-3-031-43583-6
040 $a GP $c GP
041 0 $a eng
050 4 $a QC793.5.N42
072 7 $a PHP $2 bicssc
072 7 $a SCI051000 $2 bisacsh
072 7 $a PHP $2 thema
082 0 4 $a 539.7215 $2 23
090 $a QC793.5.N42 $b S967 2023
100 1 $a Sutton, Andrew T. C. $3 1434903
245 1 0 $a Domain generalization with machine learning in the NOvA experiment $h [electronic resource] / $c by Andrew T.C. Sutton.
260 $a Cham : $c 2023. $b Springer Nature Switzerland : $b Imprint: Springer,
300 $a xi, 170 p. : $b illustrations (chiefly color), digital ; $c 24 cm.
490 1 $a Springer theses, $x 2190-5061
500 $a "Doctoral Thesis accepted by the University of Virginia, USA."
505 0 $a Chapter 1: Neutrinos: A Desperate Remedy -- Chapter 2. A Review of Neutrino Physics -- Chapter 3. The NOvA Experiment -- Chapter 4. Event Reconstruction -- Chapter 5. The 3-Flavor Analysis -- Chapter 6. A Long Short-Term Memory Neural Network -- Chapter 7. Domain Generalization by Adversarial Training -- Chapter 8. Conclusion.
520 $a This thesis presents significant advances in the use of neural networks to study the properties of neutrinos. Machine learning tools like neural networks (NN) can be used to identify the particle types or determine their energies in detectors such as those used in the NOvA neutrino experiment, which studies changes in a beam of neutrinos as it propagates approximately 800 km through the earth. NOvA relies heavily on simulations of the physics processes and the detector response; these simulations work well, but do not match the real experiment perfectly. Thus, neural networks trained on simulated datasets must include systematic uncertainties that account for possible imperfections in the simulation. This thesis presents the first application in HEP of adversarial domain generalization to a regression neural network. Applying domain generalization to problems with large systematic variations will reduce the impact of uncertainties while avoiding the risk of falsely constraining the phase space. Reducing the impact of systematic uncertainties makes NOvA analysis more robust, and improves the significance of experimental results.
650 0 $a Neutrinos. $3 786066
650 0 $a Neural networks (Computer science) $3 528588
650 1 4 $a Particle Physics. $3 1366292
650 2 4 $a Accelerator Physics. $3 1387873
650 2 4 $a Measurement Science and Instrumentation. $3 769080
650 2 4 $a Machine Learning. $3 1137723
650 2 4 $a Computational Physics and Simulations. $3 1366360
710 2 $a SpringerLink (Online service) $3 593884
773 0 $t Springer Nature eBook
830 0 $a Springer theses. $3 831604
856 4 0 $u https://doi.org/10.1007/978-3-031-43583-6
950 $a Physics and Astronomy (SpringerNature-11651)