Choose timezone
Your profile timezone:
Diboson Group
June 18, 2012
Common ATLAS/CMS LHC EWG di-boson meeting
Discussion topics:
- TGC parameter limit setting methods
- Content of ATLAS/CMS di-boson note
*** TGC's ***
Masahiro Morii and Lindsey Gray presented the TGC limit setting methods
on ATLAS and CMS. Both experiments use similar techniques to parametrize
the yield as function of anomalous coupling, namely event-by-event reweighting (ATLAS)
and grid parameterization (ATLAS,CMS) taking into account reconstruction efficiency
variations.
The statistical methods to determine the 95% CI are Log Likelihood based, nuisance
parameters are included as Gaussian variations. The following variants were discussed:
a) Baysian limits (ATLAS, CMS)
b) Profile likelihood limits (ATLAS, CMS)
c) Neyman construction (ATLAS)
d) CLS (CMS)
It was agreed to drop the CLS method d). It is not adequate for TGC limits since
the SM point is always included by construction. Currently, a) and c) are planned for
publications of ATLAS and CMS.
a) WW/Wgam b) ZZ c) WZ were used in ATLAS; for CMS: b) WW, WGamma and ZGamma.
For a possible combination of limits between ATLAS and CMS it was agreed that
the parameters to construct the likelihood (yield, nuisance parameters, correlations)
should be used. Not yet clear whether the parameters must be made public before
combination. No decision was taken whether to use method a) or c) for combination.
For 2D limits method c) might be computing intensive.
AI: Need to identify common format to exchange systematics, yields etc for loglikelihood
AI: Identify correlated systematicds between CMS and ATLAS (theory,lumi?)
The parametrization of TGC's were discussed. The choice of parameters will be documented
to have a common set for both experiments. For combination across channels less
constraints should be used (eg. don't use HISZ for WW when combining with WZ)
to make the combination more meaningful.
*** Content of note ***
1) Theoretical cross section
- MC predictions (MCFM, POWHEG-BOX, etc…)
- Theoretical uncertainties (PDF, Scale, ..)
2) Fiducial cross section
- fiducial cross section definition
- object definitions (dressed vs non dressed leptons)
- common systematics
3) TGC's
- parametrisations
- extraction and limit setting
- systematics (per experiments, and correlations between experiments)
- exchange format
Proposal for editors:
1) (Vuko)
2) (Matt/SC)
3) TGC (Alex)
No unfolding planned yet, could be added later.
Diboson Group (18 Jun 2012)
Chaired by: Oh, Alexander; Herndon, Matthew; Hsu, Shih-Chieh; Brigljevic, Vuko
There are minutes attached to this event.
Show them.