Development of novel statistical methods and application to the ATLAS experiment at the LHC

The project concerns development of novel statistical methods for data analysis, in areas such as statistical models with uncertain error parameters, measurement of differential distributions using deconvolution (unfolding), and machine learning techniques for statistical tests, multivariate regression and density estimation. The statistical tools will be applied to physics analyses using data that will be collected by the ATLAS detector at the LHC. The physics application will involve the search for new processes in one of several possible areas, such as top-quark or Higgs boson properties. The measurements will use data from the upcoming data-taking run (run 3) with proton-proton collisions at 14 TeV. In addition, the student will contribute to the ATLAS experiment through development of statistical software for the ATLAS Collaboration as well as to the operation of the ATLAS detector at CERN.