logo SBA

ETD

Archivio digitale delle tesi discusse presso l’Università di Pisa

Tesi etd-10252012-120538


Tipo di tesi
Tesi di dottorato di ricerca
Autore
SFORZA, FEDERICO
URN
etd-10252012-120538
Titolo
Evidence for Diboson Production in the Lepton plus Heavy Flavor Jets Final State at CDF
Settore scientifico disciplinare
FIS/04
Corso di studi
FISICA
Relatori
tutor Dott. Chiarelli, Giorgio
commissario Tonelli, Guido
commissario Grivaz, Jean-Francois
commissario Cobal, Marina
commissario Busetto, Giovanni
Parole chiave
  • hadron collider
  • diboson
  • CDF
  • lepton plus jets
  • SVM
Data inizio appello
29/10/2012
Consultabilità
Completa
Riassunto
The topic of this thesis is the measurement of W and Z bosons associate production in the lepton (l) plus neutrino (nu) plus Heavy Flavor (HF) quarks final state identified by the CDF II experiment at the Tevatron collider at s^1/2=1.96 TeV.

The fist evidence at hadron colliders of diboson production in such final state is reported.

The CDF II experiment set tight constraints on the existence of the Higgs boson with most of the sensitivity (for the low mass Higgs boson) coming from a very similar final state to the one studied in this thesis (WH -> l nu+b bbar). Thus it is possible to test the correctness of the analysis procedure used in the Higgs search.

The actual analysis procedure is divided in several steps: object identification, background estimate and statistical analysis.

First, a set of advanced identification algorithms is exploited for the recognition of the final state objects: one charged lepton a neutrino, and two high energy jets, of which at least one tagged by the identification of the secondary decay vertex produced by a HF hadron.

Between the specific event selection requirements an original technique, based on the support vector machine algorithm was developed to suppress the multi-jet events, a background, difficult to model, due to events in which no real W -> l nu decay is present.

Successively the total background is estimated with a variety of methods (both Monte Carlo and data driven). Finally a shape analysis of the di-jet invariant mass distribution and of a flavor-separator neural network distribution allowed the extraction of the combined diboson signal and the separate WW vs WZ/ZZ contributions.
File