Andrea Beschi

The Compact Muon Solenoid (CMS) experiment implements a sophisticated two-level triggering system composed of a custom-designed hardware Level-1 trigger and a software High-Level-Trigger.A new Level-1 trigger architecture with improved performance is now being used to maintain thresholds compatible with the CMS physics program despite the challenging luminosity conditions experienced during Run II.We present the performance of the upgraded CMS electron and photon trigger.The calorimeter trigger system plays a central role in achieving the ambitious physics program of Run II.The upgraded trigger uses the granularity of the calorimeters to optimally reconstruct the electromagnetic trigger objects.The performance of the new trigger system will be presented, based on proton-proton collision data collected in Run II.The selection techniques used to trigger efficiently will be presented, along with the strategies employed to guarantee efficient triggering for new resonances and other new physics signals involving electron/photon final states.

Questo elaborato ha come obbiettivo quello di studiare i Word Embeddings, in relazione alla loro applicazione alla collezione biomedicale di PubMed. L'interesse e quello di sviluppare una versione del modello SkipGram con la quale addestrare i Word Embeddings con la collezione PubMed sfruttando una piattaforma con poche risorse di calcolo come un Notebook: e stata sviluppata un'implementazione in Python del modello utilizzando TensorFlow e i risultati sono stati valutati in modo intrinseco.

This thesis presents the analysis of the data collected in 2016 and in 2017 by the CMS detector at the Large Hadron Collider to perform a precise measurement of the cross section of the Higgs boson production in association with a pair of top-antitop quarks (ttH). The diphoton decay channel of the Higgs boson is exploited to perform the measurement, as among the most sensitive channels. The analysis exploits 35.9 fb−1 of data collected in 2016 and 41.5 fb−1 collected in 2017 in proton-proton collisions at a centre-of-mass energy of 13 TeV.At the beginning of this work, the measured cross section of the ttH process showed a mild departure from the standard model (SM) expectation of about two standard deviations. A more precise determination was necessary to ascertain whether the observed tension was due to a statistical fluctuation or to some unpredicted feature.The choice of the diphoton decay channel is due to the fully reconstructed final state and to the low number of events arising from different processes but with similar topology, namely background. The experimental signature of a Higgs boson decaying to a pair of photons is the presence of a narrow resonant peak arising in the invariant mass distribution of the photon pairs. The ttH production can be exclusively identified by the presence in the final state of the decay products of the top quarks. The ttH production rate is extracted from a fit to the invariant mass spectrum of the photon pairs. The fit function is based on a signal model, built from simulation of the Higgs boson production processes, and a background one, derived directly from fitting the data.The fit is performed with floating the signal strength μ, defined as the ratio between the measured ttH production cross section and the SM expectation. The analysis of the 2016 data resulted in an observed value of μ = 2.2+0.9−0.8, rejecting the background-only hypothesis at 3.2 standard deviations, where 1.5 is expected for a SM Higgs boson. The refinement of the techniques exploited for the data analysis lead to an improvement in the sensitivity of more than 50 % in the analysis of the 2017 data, with μ = 1.3+0.7−0.5, rejecting the background-only hypothesis at 3.1 standard deviations, where 2.2 are expected for a SM Higgs boson. The combination of the two analyses resulted in an observed signal strength of μ = 1.7+0.6-0.5, corresponding to a significance of 4.1 standard deviations.The analysis of the 2016 data, together with the analyses targeting final states with the Higgs boson decaying in b quarks, vector bosons and τ leptons, allowed the first experimental observation of the ttH process. The result exploited the analysis of the data collected in 2016 at a centre-of-mass energy of 13 TeV, as well as the data collected in 2011 and 2012 at a centre-of-mass energy of 7 and 8 TeV, respectively. The best fit value of the combination of the different channels is μ = 1.26+0.31−0.26, in agreement with the SM expectation. The background-only hypothesis is rejected at the level of 5.1 standard deviations. This result proves for the first the tree-level coupling of the Higgs boson with the top quarks and, hence, with an up-type quark.