• norsk
    • English
  • English 
    • norsk
    • English
  • Login
View Item 
  •   Home
  • Faculty of Mathematics and Natural Sciences
  • Department of Physics and Technology
  • Department of Physics and Technology
  • View Item
  •   Home
  • Faculty of Mathematics and Natural Sciences
  • Department of Physics and Technology
  • Department of Physics and Technology
  • View Item
JavaScript is disabled for your browser. Some features of this site may not work without it.

Parameter Scans and Machine Learning for beyond Standard Model Physics

Strümke, Inga
Doctoral thesis
Thumbnail
View/Open
PDF (9.482Mb)
URI
https://hdl.handle.net/1956/20546
Date
2019-05-02
Metadata
Show full item record
Collections
  • Department of Physics and Technology [1814]
Abstract
This thesis focuses on different beyond Standard Model theories, and the use of statistical methods to investigate them. Supersymmetry is considered in different contexts. First, the supersymmetry breaking scheme of gaugino mediation is investigated. These models were previously thought to be ruled out after observation of the lightest Higgs with mass 125GeV, but we show that non-vanishing trilinear terms can provide a sufficiently large Higgs mass via mixing in the stop sector. Introducing the concept of machine learning, two models are investigated using different machine learning techniques. A two-Higgs doublet model with mass degenerate neutral Higgses is studied, and deep learning is used to separate the two CP-states whose decay products have very similar kinematics. The challenge of prior dependence associated with problems relying on simulated training data and the probabilistic interpretation of classifier output is discussed. Second, another supersymmetric scenario, this time with a sneutrino as the lightest observable supersymmetric particle, is considered from a collider perspective. A boosted decision tree is used to detect signal events in datasets with small signal mixture parameters. Later, a supersymmetric scenario with a gravitino as lightest and neutralino as next-to-lightest supersymmetric particle is considered in a cosmological context. The presence of late-decaying neutralinos can potentially come in conflict with constraints from Big Bang Nucleosynthesis, and we investigate a specific region in parameter space where resonant annihilation via a heavy Higgs lowers the neutralino relic abundance. In addition to this constraint, collider searches and the observed dark matter abundance are combined to form a likelihood which guides a scan through the parameter space. Dark matter is further investigated from a model-independent viewpoint. The potentially sharp gamma-ray features in a signal from dark matter annihilation into Standard Model excited meson states are simulated, and we show how such a signature would stand out from the astrophysical background. The relevant energy range for this would be what has come to be called the “MeV-gap”, since this range is astrophysically relatively unexplored.
Has parts
Paper I: J. Heisig, J. Kersten, N. Murphy, and I. Strümke, Trilinear-augmented gaugino mediation. Journal of High Energy Physics 2017 (2017) no. 5, 3. The article is available at: http://hdl.handle.net/1956/20545

Paper II: A. Kvellestad, S. Maeland, and I. Strümke, Signal mixture estimation for degenerate heavy Higgses using a deep neural network. The European Physical Journal C 78 (2018) no. 12, 1010. The article is available at: http://hdl.handle.net/1956/20405

Paper III: T. Bringmann, A. Hryczuk, A. Raklev, I. Strümke, and J. Van den Abeele, Smoking gun dark matter signatures in the MeV range, 52-53. In: A. D. Angelis, et al., Science with e-ASTROGAM: A space mission for MeV - GeV gamma-ray astrophysics. Journal of High Energy Astrophysics (2018) no. 19, 1-106. The article is available in the main thesis. The article is also available at: https://doi.org/10.1016/j.jheap.2018.07.001
Publisher
The University of Bergen
Copyright
Copyright the author. All rights reserved

Contact Us | Send Feedback

Privacy policy
DSpace software copyright © 2002-2019  DuraSpace

Service from  Unit
 

 

Browse

ArchiveCommunities & CollectionsBy Issue DateAuthorsTitlesSubjectsDocument TypesJournalsThis CollectionBy Issue DateAuthorsTitlesSubjectsDocument TypesJournals

My Account

Login

Statistics

View Usage Statistics

Contact Us | Send Feedback

Privacy policy
DSpace software copyright © 2002-2019  DuraSpace

Service from  Unit