indico of the meeting:
https://indico.cern.ch/conferenceDisplay.py?confId=231636
CMS talk (M. Malberti):
------------------
- Content:
a set of plots which show the effect of the MVA selection on dijet
e.g. basic jet quantities + ptjj, deltaphi(jj,gg), pt(jjgg)
- Questions and comments:
* what powheg? R: normal powheg, no minlo
* please redo the plot for ggH. R: done and added into the new
version of the talk
* how do you prevent to not have deltaphi(gg,jj) values close to pi?
In atlas we impose that MVA do not use deltaphi(gg,jj) variables
above a given threshold. That make efficiency in that region ~flat.
R: we don't. We will check the efficiency behavior.
1) ATLAS talk (D. Gillberg):
---------------------
- Content
a set of plots which show the effect of the MVA selection on dijet
e.g. basic jet quantities + ptjj, deltaphi(jj,gg), pt(jjgg)
- Question and comments:
* why do you evaluate systematics at 95% efficiency point? R:
arbitrary choice, only to get a cross-check what to expect from method
shown in later in the other talk
* would be useful to have efficiency plots for both CMS and ATLAS
* imposing that MVA does not use deltaphi(gg,jj) above a given
threshold reduces separation power. Maybe a trade-off
between loss in separation and theory uncertainty would
be beneficial. R: not for now but interesting for future studies
2) ATLAS talk (F. U. Bernlochner):
--------------------------
- Content
development of a new technique to evaluate systematics due to
radiation of extra jets in ggH for dijet channel when using MVAs.
Currently ST-like uncertainties used as input (technique itself can
be used with any input uncertainties)
- Question and comments:
* how do you deal with MCFM for regions with negative cross
sections? R: Currently, no shape information is used below a certain
cut but only normalization. (In principle we can extend the method in
that region but once MCFM-only uncertainties are > 100%
it becomes difficult to interpret the result.)
* are there more checks that the method works, even if it uses
categorization based on a single variable? page 11 looks like a
trivial closure test. R: it is the test at page 9
* why is the uncertainty in page 9 different between pure MCFM
and powheg reweighted? R: spectra in the two MCs are different, we
were actually happy to see this agreement
* why not doing it on the MVA directly, i.e. by applying MVA to MCFM
ntuple output more than via deltaphi method?
R: technically more involved, maybe using MVA directly on MCFM would
be important to understand if it makes a difference if the uncertainty
is 35% vs. 40% vs. 45% but given the present statistical
uncertainty of this channel would not make such a difference
* is it possible to get weights from ATLAS and apply them to CMS
montecarlo? R: good idea to have consistency.
Weights could be ready by the beginning of next week.
ACTION ITEMS:
------------
* Experiments should discuss on how to proceed to have a common
approach and in particular have numbers which can be directly compared.
* The technique developed by Florian and Dag (ATLAS) and using inputs
from F. Tackmann et al. paper can be used by CMS physicists
* There is the Issue that ATLAS is imposing flat efficiency above a
given threshold in deltaphi and the deltaphi bins are based on
that. CMS can check the efficiency above that value.
* Maybe the starting point is to apply it to the Hgg channel with deltaphi
as baseline variable
* Once enough experience other channels can follow (using the
appropriate baseline variable for them, e.g. pTHjj)
* Code from ATLAS can be ready by beginning of next week and passed
to the LHC working group which will make it public (with
documentation to follow)