Beyond cuts in small signal scenarios

We investigate enhancing the sensitivity of new physics searches at the LHC by machine learning in the case of background dominance and a high degree of overlap between the observables for signal and background. We use two different models, XGBoost and a deep neural network, to exploit correlations...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:The European physical journal. C, Particles and fields Ročník 83; číslo 5; s. 379 - 22
Hlavní autoři: Alvestad, Daniel, Fomin, Nikolai, Kersten, Jörn, Maeland, Steffen, Strümke, Inga
Médium: Journal Article
Jazyk:angličtina
Vydáno: Heidelberg Springer 01.05.2023
Springer Nature B.V
SpringerOpen
Témata:
Analysis
Artificial neural networks
Decision trees
Kinematics
Large Hadron Collider
Machine learning
Neural networks
Physics
Sensitivity enhancement
ISSN:1434-6044, 1434-6052
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:We investigate enhancing the sensitivity of new physics searches at the LHC by machine learning in the case of background dominance and a high degree of overlap between the observables for signal and background. We use two different models, XGBoost and a deep neural network, to exploit correlations between observables and compare this approach to the traditional cut-and-count method. We consider different methods to analyze the models' output, finding that a template fit generally performs better than a simple cut. By means of a Shapley decomposition, we gain additional insight into the relationship between event kinematics and the machine learning model output. We consider a supersymmetric scenario with a metastable sneutrino as a concrete example, but the methodology can be applied to a much wider class of models.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1434-6044
1434-6052
DOI:10.1140/epjc/s10052-023-11532-9