Learning to rank Higgs boson candidates

Abstract

AbstractIn the extensive search for new physics, the precise measurement of the Higgs boson continues to play an important role. To this end, machine learning techniques have been recently applied to processes like the Higgs production via vector-boson fusion. In this paper, we propose to use algorithms for learning to rank, i.e., to rank events into a sorting order, first signal, then background, instead of algorithms for the classification into two classes, for this task. The fact that training is then performed on pairwise comparisons of signal and background events can effectively increase the amount of training data due to the quadratic number of possible combinations. This makes it robust to unbalanced data set scenarios and can improve the overall performance compared to pointwise models like the state-of-the-art boosted decision tree approach. In this work we compare our pairwise neural network algorithm, which is a combination of a convolutional neural network and the DirectRanker, with convolutional neural networks, multilayer perceptrons or boosted decision trees, which are commonly used algorithms in multiple Higgs production channels. Furthermore, we use so-called transfer learning techniques to improve overall performance on different data types.

Autoren

Marius Köppel
Alexander Segner
Martin Wagener
Lukas Pensel
Andreas Karwath
Christian Schmitt
Stefan Kramer

DOI

10.1038/s41598-022-10383-w

eISSN

2045-2322

Ausgabe der Veröffentlichung

1

Zeitschrift

Scientific Reports

Sprache

en

Artikelnummer

13094

Online publication date

2022

Status

Published online

Herausgeber

Springer Science and Business Media LLC

Herausgeber URL

http://dx.doi.org/10.1038/s41598-022-10383-w

Datum der Datenerfassung

2022

Titel

Learning to rank Higgs boson candidates

Ausgabe der Zeitschrift

12

Datenquelle: Crossref

Abstract

In the extensive search for new physics, the precise measurement of the Higgs boson continues to play an important role. To this end, machine learning techniques have been recently applied to processes like the Higgs production via vector-boson fusion. In this paper, we propose to use algorithms for learning to rank, i.e., to rank events into a sorting order, first signal, then background, instead of algorithms for the classification into two classes, for this task. The fact that training is then performed on pairwise comparisons of signal and background events can effectively increase the amount of training data due to the quadratic number of possible combinations. This makes it robust to unbalanced data set scenarios and can improve the overall performance compared to pointwise models like the state-of-the-art boosted decision tree approach. In this work we compare our pairwise neural network algorithm, which is a combination of a convolutional neural network and the DirectRanker, with convolutional neural networks, multilayer perceptrons or boosted decision trees, which are commonly used algorithms in multiple Higgs production channels. Furthermore, we use so-called transfer learning techniques to improve overall performance on different data types.

Addresses

Johannes Gutenberg University, Mainz, Germany. mkoeppel@uni-mainz.de.

Autoren

Marius Köppel
Alexander Segner
Martin Wagener
Lukas Pensel
Andreas Karwath
Christian Schmitt
Stefan Kramer

DOI

10.1038/s41598-022-10383-w

eISSN

2045-2322

Externe Identifier

PubMed Identifier: 35908043
PubMed Central ID: PMC9338962

Funding acknowledgements

Medical Research Charities Group: MR/S003991/1
Johannes Gutenberg-Universität Mainz:
Medical Research Council: MR/S003991/1

Open access

true

ISSN

2045-2322

Ausgabe der Veröffentlichung

1

Zeitschrift

Scientific reports

Sprache

eng

Medium

Electronic

Online publication date

2022

Open access status

Open Access

Paginierung

13094

Datum der Veröffentlichung

2022

Status

Published

Publisher licence

CC BY

Datum der Datenerfassung

2022

Titel

Learning to rank Higgs boson candidates.

Sub types

research-article
Journal Article

Ausgabe der Zeitschrift

12

Files

https://www.nature.com/articles/s41598-022-10383-w.pdf https://europepmc.org/articles/PMC9338962?pdf=render

Datenquelle: Europe PubMed Central

Abstract

In the extensive search for new physics, the precise measurement of the Higgs boson continues to play an important role. To this end, machine learning techniques have been recently applied to processes like the Higgs production via vector-boson fusion. In this paper, we propose to use algorithms for learning to rank, i.e., to rank events into a sorting order, first signal, then background, instead of algorithms for the classification into two classes, for this task. The fact that training is then performed on pairwise comparisons of signal and background events can effectively increase the amount of training data due to the quadratic number of possible combinations. This makes it robust to unbalanced data set scenarios and can improve the overall performance compared to pointwise models like the state-of-the-art boosted decision tree approach. In this work we compare our pairwise neural network algorithm, which is a combination of a convolutional neural network and the DirectRanker, with convolutional neural networks, multilayer perceptrons or boosted decision trees, which are commonly used algorithms in multiple Higgs production channels. Furthermore, we use so-called transfer learning techniques to improve overall performance on different data types.

Date of acceptance

2022

Autoren

Marius Köppel
Alexander Segner
Martin Wagener
Lukas Pensel
Andreas Karwath
Christian Schmitt
Stefan Kramer

Autoren-URL

https://www.ncbi.nlm.nih.gov/pubmed/35908043

DOI

10.1038/s41598-022-10383-w

eISSN

2045-2322

Externe Identifier

PubMed Central ID: PMC9338962

Funding acknowledgements

Medical Research Council: MR/S003991/1
Medical Research Charities Group: MR/S003991/1

Ausgabe der Veröffentlichung

1

Zeitschrift

Sci Rep

Sprache

eng

Country

England

Paginierung

13094

PII

10.1038/s41598-022-10383-w

Datum der Veröffentlichung

2022

Status

Published online

Titel

Learning to rank Higgs boson candidates.

Sub types

Journal Article

Ausgabe der Zeitschrift

12

Datenquelle: PubMed

Author's licence

CC-BY

Autoren

Marius Köppel
Alexander Segner
Martin Wagener
Lukas Pensel
Andreas Karwath
Christian Schmitt
Stefan Kramer

Hosting institution

Universitätsbibliothek Mainz

Sammlungen

DFG-491381577-G

Resource version

Published version

DOI

10.1038/s41598-022-10383-w

Funding acknowledgements

Gefördert durch die Deutsche Forschungsgemeinschaft (DFG) - Projektnummer 491381577

File(s) embargoed

false

Open access

true

ISSN

2045-2322

Zeitschrift

Scientific reports

Schlüsselwörter

530 Physik
530 Physics

Sprache

eng

Open access status

Open Access

Paginierung

13094

Datum der Veröffentlichung

2022

Public URL

https://openscience.ub.uni-mainz.de/handle/20.500.12030/8454

Herausgeber

Springer Nature

Datum der Datenerfassung

2022

Datum, an dem der Datensatz öffentlich gemacht wurde

2022

Zugang

Public

Titel

Learning to rank Higgs boson candidates

Ausgabe der Zeitschrift

12

Files

learning_to_rank_higgs_boson_-20221129142953611.pdf

Datenquelle: OPENSCIENCE.UB

Files

Files

Werkzeuge