Bonjour

je n'arrive pas à envoyer sur la liste de diffusion du LIPN

---------- Forwarded message ----------
From: Mail Delivery System < MAILER-DAEMON@lipn.univ-paris13.fr>
Date: 20 Nov 2007 17:08
Subject: Undelivered Mail Returned to Sender
To: lebbah@gmail.com

This is the mail system at host mail.lipn.univ-paris13.fr.

I'm sorry to have to inform you that your message could not
be delivered to one or more recipients. It's attached below.

For further assistance, please send mail to postmaster.

If you do so, please include this problem report. You can
delete your own text from the attached returned message.

                  The mail system

<tartenpion@lipn.univ-paris13.fr> (expanded from
   < lipn.tous@lipn.univ-paris13.fr>): unknown user: "tartenpion"

Final-Recipient: rfc822; tartenpion@lipn.univ-paris13.fr
Original-Recipient: rfc822;lipn.tous@lipn.univ-paris13.fr
Action: failed
Status: 5.1.1
Diagnostic-Code: X-sendmail-8-9-3/8-9-2; unknown user: "tartenpion"


---------- Forwarded message ----------
From: "Mustapha Lebbah" < Mustapha.Lebbah@lipn.univ-paris13.fr>
To: lipn.tous@lipn.univ-paris13.fr
Date: Tue, 20 Nov 2007 17:06:13 +0100
Subject: Séminaire Apprentissage, Jeudi 22 Nov à 13h



Bonjour à tous

Dans le cadre du séminaire Apprentissage AAPN : Apprentissage Artificiel à
Paris Nord, nous accueillerons Mr Sandro Saitta, doctorant au département
informatique, à l'Ecole Polytechnique Fédérale de Lausanne (EPFL). Il aura
l'occasion de nous faire un séminaire intitulé : Cluster Validity and
Feature Selection

Ce séminaire se tiendra le jeudi 22 novembre à 13h en salle B311 dans les
locaux du LIPN.

Merci de diffuser cette annonce.

Pour l'accès
http://www-lipn.univ-paris13.fr/planfac/?lang=fr
http://www-lipn.univ-paris13.fr/A3/Seminaire/

Bien cordialement
Mustapha Lebbah

Cluster Validity and Feature Selection
Sandro Saitta, PhD student in Computer Science
IMAC, Ecole Polytechnique Fédérale de Lausanne (EPFL), Switzerland

Estimating the number of clusters in a data set is an important task of
clustering. For example, algorithms such as K-means start with the number
of clusters as an input. We propose a new bounded cluster validity index
that is computationally efficient and can detect the single cluster case.
The new index is found to be better than several indices that have been
proposed in the literature. Feature selection reduces the dimensionality
of a data set in order, for example, to better understand the data.
Existing methods such as Genetic Algorithm based techniques, although
efficient, need complex tuning of several search parameters. We will
introduce a wrapper approach combining a global search algorithm (PGSL)
with the Support Vector Machines (SVM). Results show that the new method
is efficient and selects a small number of features for the benchmark data
sets.


-------------------------------------------
Laboratoire LIM&BIO, Université Paris Nord
http://www.limbio-paris13.org
http://lebbah.googlepages.com/home
-------------------------------------------





--
-------------------------------------------------------------
Laboratoire LIM&BIO, Université Paris Nord
http://www.limbio-paris13.org
http://lebbah.googlepages.com/
--------------------------------------------------------------