Join the disputation
The PhD defence and trial lecture will be hybrid. the attendees can decide whether they want to attend in-person or via Zoom Webinar. The room opens for participation just before the disputation starts, and closes for new participants approximately 15 minutes after the defence has begun.
Join the trial lecture - 18th of September at 10:15 (Aud. 209, Svein Rosselands Hus/Zoom)
"Applications of machine learning in cosmology -- promise and perils"
Conferral summary
Standardmodellen for Universet inneholder flere ingredienser som er lite forstått i partikkelfysikken. Flere alternative modeller som belyser deres egenskaper, kan samsvare med empirisk data, og må utforskes ved bruk av numeriske simuleringer. Dette arbeidet har utforsket en klasse modeller som gjennomgår en faseovergang i de senere stadiene av universets utvikling, og har funnet relevante signaturer i strukturdannelse og gravitasjonsbølger.
Main research findings:
"The dark sector" of cosmology refers to exotic dark components that modify the gravitational dynamics in the Universe to align with empirical observations. These dark components are not unique in producing the observed behaviour, and several alternative models will produce additional effects that can be used to distinguish them in upcoming surveys. This work explores in detail a class of models that brings the Universe through a dark phase transition, that produces dark topological defects, gravitational waves and boosts structure formation.
To better understand these models, I develop simulation software that captures the relevant gravitational dynamics and structure formation. The software is applied to find clear signatures of the alternative ingredients, which can be detected in upcoming experiments, such as imprints on the cosmic microwave background, boosted structure formation, and production of gravitational waves. I then continue to propose and simulate a novel and complex mechanism in which this class of models can produce the observed late-time cosmic acceleration. Such models would have far-reaching implications for both the origins and the ultimate future of the Universe
