Validation of procedures used by CMS in the characterization of Higgs candidate events.

Abstract

With the recent announcement of the discovery of a new bosonic state at the Large Hadron Collider (LHC), hypothesized to be the long-sought Higgs boson, the correctness of the procedures used for characterization and classification of this state has become very relevant. This project worked on a number of different aspects in the analysis of the HZZ4l decay channel, related especially to lepton detection. Characterization of this channel relies on the accurate prediction of lepton detection efficiency and resolution, but currently used algorithms for prediction of these detector effects are computationally demanding. This project developed a validation procedure for a faster simulation algorithm written by the Caltech group, and showed its adequacy for predicting efficiency and resolution. Another issue analyzed in lepton detection was the use of multivariate analysis techniques, such as boosted decision trees, for reconstruction of the actual lepton energies from detected values, in order to discover whether the current reconstruction algorithms can be improved. Some questions outside of the scope of lepton detection were also analyzed, such as the distributions of the two intermediate Z in the HZZ4l decay. The official analysis uses an invariant mass proximity criterion to perform the pairing, but there are other possible pairing criteria that could be helpful in the detection of new physics. Current data in the signal region is scarce, but the algorithms developed for this pairing analysis will be available for future, larger datasets.

Additional details