Reconstruction of PINGU data with a Differential Evolution Minimizer
Student paper other, 20 credits / 30 HE creditsStudent thesis
The Precision IceCube Next Generation Upgrade (PINGU) is supposed to have an energy threshold below 10 in order to resolve the neutrino mass hierarchy. In order to reconstruct the energy and direction of neutrinos interacting in this array, producing both a hadronic cascade and a muon track, advanced reconstruction methods need to be employed. A class of these seeks to maximize a complicated likelihood function within an 8-dimensional parameter space describing the event, and requires sophisticated minimizers to achieve the necessary resolution in a reasonable time. In this report, a pre-existing but hitherto unused minimizer which samples that parameter space with several Markov chains at once, based on the Differential Evolution Monte Carlo algorithm, is developed further and its behaviour and performance is tested on simulated data of the IceCube/PINGU array. The tests compare both various configurations of the minimizer and Markov Chain Monte Carlo, a similar previous approach.
Place, publisher, year, edition, pages
2013. , 46 p.
Other Physics Topics
IdentifiersURN: urn:nbn:se:uu:diva-214617OAI: oai:DiVA.org:uu-214617DiVA: diva2:685277
Pérez de los Heros, Carlos, Associate Professor
Hallgren, Allan, Professor