Discriminación de eventos de producción de pares de quarks top del Modelo Estándar, de la producción de materia oscura en asociación con un par de quark tops utilizando una red neuronal artificial

ABSTRACT: The Standard Model of particles condenses our knowledge about the fundamental blocks of matter and their interactions. In spite of the great success of this theory describing the experiments up to date, there are several shortcomings suggesting that this might not be the ultimate particle physics theory. One of these shortcomings is related to the fact that none of the particles considered in the Standard Model can account for the so called Dark Matter, an unknown component of the Universe which interacts gravitationaly as the ordinary matter but apparently does not interact in any other way via the electromagnetic, weak or strong forces. Searches for Dark Matter are happening worldwide in different kinds of experiments and/or observatories. In particular, Dark Matter could be produced in particle colliders such as the Large Hadron Collider at the European Center for Particle Physics (CERN). These experiments produce huge amounts of data that have to be carefully analyzed to find indications of new particles in a huge ocean of known, background-like, collisions. In order to help on this task, machine learning algorithms have been extensively used in particle physics data analysis since many years ago to perform classification, discrimination, reconstruction or pattern recognition in a large variety of analysis. This work frames in the context of a search for Dark Matter particles in association with a top quark-antiquark pair in the Compact Muon Solenoid (CMS) experiment at the LHC. Simulated 3 Monte Carlo data of proton-proton collisions at a center of mass energy of 13 TeV are used to train an Artificial Neural Network to discriminate between the signal model: dark matter in association with a top quark pair, and the background: standard model production of top quark pairs. The work is divided in several chapters. Chapter 2 provides a general introduction to the standard model of particle physics, the dark matter and also to the Large Hadron Collider and the CMS experiment. Chapter 3 presents a short overview of the Machine Learning algorithms with special emphasis on the Artificial Neural Networks. Chapter 4 presents an explanation of the most important kinematic variables used for this kind of analysis and also the way in which an Artificial Neural Network has been implemented, optimized and trained. Chapter 5 presents the final results and finally, chapter 6 shows the conclusions. technique has been implemented into the MAST tool, developed by the ISTR group of the UC. Therefore, a benchmark battery for MAST will be proposed in order to contrast the results of the new analysis technique with the existing one.technique has been implemented into the MAST tool, developed by the ISTR group of the UC. Therefore, a benchmark battery for MAST will be proposed in order to contrast the results of the new analysis technique with the existing one.Máster en Física de Partículas y del Cosmo