[Nnhep] Proposal for the meeting

[Lev Dudko]

    Hello Tomas,
 What time is appropriate for you? I will have shift 28.05 till 16.30,
after that I am able to meet. On Friday I will have meeting at 9.30 till
11 (11.30) and after that I can meet in any time appropriate for you. It
is also possible to meet on Saturday.
       Best regards,
            Lev

В Срд, 27/05/2009 в 07:31 +0100, Tomas Lastovicka пишет:
> 
> Dear all,
> 
> i am traveling to CERN today from Montenegro. Unfortunately, i will be
> busy with the CLIC workshop on both Thursday and Friday the whole day
> and thus i am not sure i could join you at 29.05.09 11.00-12.00. I
> will contact you once at CERN.
> 
> Best,
> 
> Tomas
> 
> 
> 
> -----Original Message-----
> From: nnhep-bounces at projects.hepforge.org on behalf of Lev Dudko
> Sent: Tue 5/26/2009 12:02 PM
> To: nnhep at projects.hepforge.org
> Cc: Konstantin.Toms at cern.ch
> Subject: [Nnhep] Proposal for the meeting
> 
> 
>      Dear Colleagues,
>   we have discussed privately, the possible time of the first meeting
> dedicated to NNHEP proposal. Looks like 3 people are in CERN now.
> Sasha
> can be connected with skype.
>     The proposed time is 29.05.09 11.00-12.00  If you want to change
> the
> time, lets discuss it.
> 
>     The first version of the questions to discuss is the following:
> 
> 1. We have to define the structure of interfaces between different
> modules and understand what should be the input of the module and the
> output. 
>    Let me remind, what I mean the module, this is one independent
> component e.g. one neural network or some other algorithm which has
> some
> input and some output, internal set of the parameters and some way of
> graphical representation of what is happened with the input in this
> module.
> 
> 2. We can start to discuss what modules (algorithms) we would like to
> see as the components of the system and what parts we can take from
> the
> available external code.
> The first set could be the following:
> Feed-Forward Neural Network, Bayesian Neural Network, Boosting
> algorithm, Decision trees, Self Organizing Maps, Likelihood
> functions, ...
>       Some of the modules are realized already in the exist packages
> (e.g. Feed-Forward NN training with perfect optimizations and training
> methods are realized in MLPfit package; BNN is realised in RuleFit,
> etc.)
> 
> 3. Practical realization of the discussed plans and how we plan to
> proceed with the code development.
> 
>         Please, correct/add the set of questions above and provide
> your
> opinion.
>               Best regards,
>                     Lev
> 
> 
> 
> --
> 
>  Lev V. Dudko            e-mail:dudko at fnal.gov
>  t. +7(495)9395881      http://top.sinp.msu.ru
> 
> 
> 
> 
-- 

 Lev V. Dudko            e-mail:dudko at fnal.gov
 t. +7(495)9395881      http://top.sinp.msu.ru

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: =?koi8-r?Q?=FC=D4=C1?= =?koi8-r?Q?_=DE=C1=D3=D4=D8?=
 =?koi8-r?Q?_=D3=CF=CF=C2=DD=C5=CE=C9=D1?=
 =?koi8-r?Q?_=D0=CF=C4=D0=C9=D3=C1=CE=C1?=
 =?koi8-r?Q?_=C3=C9=C6=D2=CF=D7=CF=CA?=
 =?koi8-r?Q?_=D0=CF=C4=D0=C9=D3=D8=C0?=
Url : http://www.hepforge.org/lists-archive/nnhep/attachments/20090527/27963710/attachment.pgp