Dr Marcella Bona Project Abstracts
BSc Projects/MSci Review Projects/MSci Research/Investigative Projects
Inputs to the Unitarity Triangle fit.
To completely characterise the Standard Model, we need to extract the values of some parameters from various high energy experiments and theoretical calculation. The UTfit project is a global fit analysis that using the Bayesian statistics extracts SM parameters and SM predictions on various observables using the most updated results from experiments and lattice QCD. Some of the inputs need special statistical treatments and analyses that have to be regularly updated.
C++ programming is necessary.
The analysis is performed within the Root package (root.cern.ch) that should be installed on the used computer.
Invariant Mass Fits For The Search Of Rare B Decays Into Two Muons.
The B mesons can decay into two muons and the probability of this decay is very low but accurately calculated within the Standard Model. ATLAS experiment at LHC is searching for these decays and the crucial part of the analysis is calculating the invariant mass of any pair of two muons in the collision events and then perform a fit to the invariant mass distribution taking into account all the signal and background components.
C++ programming is necessary.
The analysis is performed within the Root package (root.cern.ch) that should be installed on the used computer.
Multivariate analysis methods for rare B decays
As the B mesons decay into two muons with a very low probability, it is fundamental to develop methods to distinguish between the actual signal and the random background component that is dominant in the collision data. There are several methods that can be used to exploit the different topologies and event characteristics and obtaining a single variable that can be used to separate between signal and background. Fisher discriminant, Boosted Decision Tree, Neural Network have to be tested and the optimal method selected for the purpose of the search of rare B decays. C++ programming is necessary.
The analysis is performed within the Root package (root.cern.ch) that should be installed on the used computer.
Multivariate analysis methods for Dark Matter searches
In the context of Dark Matter searches, it is essential to develop methods to distinguish between the predicted signal and the background component coming from known Standard Model processes. There are several methods that can be used to exploit the different topologies and event characteristics and obtaining a single variable that can be used to separate between signal and background. Fisher discriminant, Boosted Decision Tree, Neural Network have to be tested and the optimal method selected for the purpose of the search of Dark Matter in association with heavy-flavour quarks. C++ programming is necessary. The analysis is performed within the Root package (root.cern.ch) that should be installed on the used computer.
