Preface

This document is a summary of the main projects that I have carried out within my PhD in Physics at the University of Padua (UNIPD), between December 2015 and the December 2018. The main research focus, connecting the projects presented here, has been the development and application of new statistical learning techniques in particle collider experiments. Given the interdisciplinary nature of the topics discussed in this thesis, an effort has been made to discuss the research issues and solutions in a domain generic manner, so the links with the fields of statistics and machine learning are more evident. The price of such attempt has likely been a less cohesive narrative, yet I believe that is has been worth the cost for a different take on the data analysis problems at particle colliders.

Most of the work included in this report has been carried out while employed by the INFN - Sezione di Padova as an Early Stage Researcher of the AMVA4NewPhysics MSCA-ITN. AMVA4NewPhysics is a European research network (EU Horizon 2020 Grant Agreement 675440) that provided the funding and context for the ventures described in this document. Part of the results presented here were joint work with other collaborators at CMS experiment at LHC, which is based at the European Organisation for Nuclear Research (CERN).