515664 research outputs found
Sort by
TROTA performance
No full textIdentifying boosted hadronic top quarks poses a significant challenge within the CMS physics program, particularly in Standard Model measurements and searches for new phenomena. There are many excellent tools available for identifying wide-angle jets with top quark flavor.
However, to enhance reconstruction and selection efficiencies for signal events including top quarks, an approach extending beyond large radius jets is necessary. From a physics standpoint, this necessity stems from the fact that the three quarks hadronize separately, resulting in lower momentum for the top quark and a greater spread of the three resulting jets within the detector. Consequently, these jets are clustered into separate objects. In the most extreme case, the anti-kT algorithm clusters the three quarks into three distinct small radius jets. However, there is also an intermediate regime in which the number of jets can vary. For example, one large radius jet may contain the products of the W boson, while one small radius jet may represent the b quark. This regime is not well-defined due to the jet clustering algorithm's sensitivity to performance. In this work, we propose an approach to include the top quark from the fully resolved to the intermediate regime. This represents one of the first attempt to reconstruct Top quark in a wide pT range using standard jets. The algorithm is based on neural networks and utilizes information from jet kinematics and ParticleNet taggers. This approach is referred to as Top Reconstruction: an Object Tagger Algorithm (TROTA).
Its performance was studied using simulations of the 2018 data-taking period of the CMS experiment
An Energy Correlation Function Tagger for Gluon-Gluon Resonances
No full textThis note presents a tagging method for the discrimination of processes with two final state gluons from the dominant QCD background. The tagging model is a boosted decision tree that uses energy correlation functions as input features. Energy correlation functions are jet substructure observables that use the kinematic information of the constituent particles of a jet as a probe for jet features. Development of the gluon-gluon tagger makes particular use of the high-order 5-point correlation functions. The tagger is trained using data and simulated samples corresponding to the 2017 data-taking period of Run 2. The following discussion includes the process of tagger development, including the creation of a gluon-enriched control region, selection of energy correlation functions as input features, and a proposed approach for calibration. Performance of this tagging technique on Run 2 simulation from 2017 is demonstrated
Visit by Ms Antoņina Ņenaševa, Deputy Speaker of the Saeima (Latvian Parliament), Latvia
No full textVisit by Ms Antoņina Ņenaševa, Deputy Speaker of the Saeima (Latvian Parliament), Latvi
Data-parallel leading-order event generation in MadGraph5_aMC@NLO
No full textThe CUDACPP plugin for MadGraph5_aMC@NLO aims to accelerate leading order tree-level event generation by providing the MadEvent event generator with data-parallel helicity amplitudes. These amplitudes are written in templated C++ and CUDA, allowing them to be compiled for CPUs supporting SSE4, AVX2, and AVX-512 instruction sets as well as CUDA- and HIP-enabled GPUs. Using SIMD instruction sets, CUDACPP-generated amplitude routines routines are shown to speed up linearly with SIMD register size, and GPU offloading is shown to provide acceleration beyond that of SIMD instructions. Additionally, the resulting speed-up in event generation perfectly aligns with predictions from measured runtime fractions spent in amplitude routines, and proper GPU utilisation can speed up high-multiplicity QCD processes by an order of magnitude when compared to optimal CPU usage in server-grade CPUs
Energy-energy correlators in small and large systems
No full textEnergy-energy correlators (EECs) provide a powerful tool to study the evolution of scattered partons into final-state hadrons. In these proceedings, a variety of energy correlator measurements performed by the ALICE collaboration are reported. The 2-point energy-energy correlator (EEC) is measured in inclusive jets and heavy-flavor jets in pp collisions, as well as in inclusive jets in p-Pb collisions. The 3-point energy correlator is also discussed, including prospects for its use in extracting the strong coupling constant
Introduction to Quantum Computing, Quantum Machine Learning and Optimization
No full textThis talk starts with an introduction to the fundamental concepts of quantum mechanics and quantum computing. We then gain a basic understanding of quantum algorithms by exploring Deutsch and Grover's algorithms. Building on this, we will explore the key concepts of quantum machine learning (QML). The embedding of classical data and parameter optimisation methods as part of the general data processing pipeline for quantum networks is discussed in the context of parametrised quantum circuits. The presentation concludes with a consideration of the possible advantages and challenges in the QML domain, and with examples of CERN-specific use cases.
Please note that pictures and videos might be taken during the event. The pictures and videos might be used for communication about the event. By joining the lecture, you are agreeing to being featured in these communication actions. </p
Searches for Higgs Boson Decays into Dark Matter Particles in the ATLAS Experiment
No full textThis paper presents an overview of the latest results and research methodology of statistical combination of Higgs invisible searches at ATLAS Large Hadron Collider experiment using data collected in Run I ( = 7 TeV, 8 TeV) and Run II ( = 13 TeV). In this search, multiple production modes of the Standard Model Higgs boson were considered. Obtained upper limit on H → inv branching ratio of 0.107 (0.077) at the 95% confidence level is observed (expected). This result is the most strict up to date. Obtained values at the ATLAS experiment are compared with the results of direct-search experiments
Recoil-Safe Subtraction, Matching and Merging in e+e- to hadrons
No full textWe present the first next-to-leading order matched and multi-jet merged predictions based on the Alaric parton shower. The components needed for infrared subtraction in the S-MC@NLO algorithm are computed analytically for the case of color singlet decays to hadronic final states and validated against existing approaches for up to e+e- to 5 jets. Phenomenological results for e+e- to hadrons at the Z pole are obtained with up to five jets at next-to-leading order precision, for the first time using an evolution algorithm with NLL-preserving kinematics mapping