Minutes of the Terascale 'Statistical Tools group'
DESY internal Kick-off Meeting 8.5.2008:
(For details see the individual talk contributions)
There were about 20 attendees from DESY and Uni
Hamburg.
1. We think that the best way for our group to be successful
would be to make substantial contributions to the
ATLAS/CMS common software pools, e.g. in RooStats.
One possible area to make a contribution is in likelihood fits,
where, contrary to popular belief, we think that for certain
problems Minuit is not optimal or at least it would be good to
have some alternative algorithms at hand. As an example there
is the lvmini program from Volker Blobel
(see page 8 of his talk).
2. The 'user community' represented in the meeting by
Christian Sander (CMS) and Karsten Koeneke (ATLAS)
expressed a strong wish that we should provide an information
forum for questions from students to be answered by experts.
This lead us directly to the conclusion that a hypernews
or equivalent forum could be useful, possible with separate
and common areas for ATLAS and CMS.
The idea of such a forum is to have a low threshold for anserwing
questions from students, perhaps lower than the level of questions being
discussed in similar forums at CERN.
3. The user community also asked to provide regular
seminars and education.
For instance a dedicated seminar or extended 'tools meeting'
on how to deal with systematic errors
could be quite useful.
Now we think that this again should be adjusted with what
you do and provide at CERN and elsewhere - on the other
hand an overlap wouldn't hurt too much.
So we plan to have regular meetings on thursday morning 10h30,
once every two to four weeks, sometimes with outside
contributions and a focus on a special topic. Further details will be announced
in due time.
Next main events in 2008 are:
19.6. Miniworkshop on statistical tools at DESY
all members from the alliance are invited cordially!
29.9.-2.10. School on statistical tools at DESY
There are minutes attached to this event.
Show them.
