This is the first announcement of the AIO course organized by SNN Nijmegen
and SNN Amsterdam.

Advanced Issues in Neurocomputing

INTENSIVE, 4-DAYS COURSE WITH COMPUTERPRACTICAL

2 t/m 5 April 2001

ADVANCED ISSUES IN NEUROCOMPUTING
Description:
A recent development in neurocomputing is the description of learning and
reasoning in terms of probabilistic models. A similar development can also
be seen in artificial intelligence (`Uncertainty in AI') and in Bayesian
statistics. The probabilistic approach now allows us to combine adaptive
methods with expert knowledge in a robust manner, and to quantify the
reliability of the model. 
The course presents a number of new computer science methods for data
analysis (modelling, classification, data-mining) based on probabilistic
models. The subjects to be treated are grouped around ``graphical models''
and ``mixture models''. In a graphical model variables are represented by
knots, and their dependencies are represented by connections in a graph. An
example hereof is a Bayesian network: a directed graph, which defines a
multidimensional probability density function. So called ``mixture models''
also present a representation of a multidimensional probability density
function, yet in a different way, i.e. in the form of a mixture of
multivariate normal distributions.
The program is organised by the research school ASCI, in co-operation with
the Foundation for Neural Networks. The course comprises a number of
tutorials and an illustration of the subject matter by means of a computer
workshop. 
The course is aimed at PhD students with a sufficient knowledge of the
field of neural networks or statistical pattern recognition. Other
interested parties may, for a modest fee, join the course or a part of it. 
Some of the aspects of the content of this course will be the same as two
years ago. The computer practical will be completely revised.


TEMPORARY PROGRAMM

Monday 2 April: Nijmegen
09.30 Coffee
09.45-12.30 Bert Kappen and Wim Wiegerinck: SNN 
Introduction to graphical models and Bayesian statistics
12.30-13.30     Lunch
13.30-16.00 Computer practical

Tuesday 3 April: Nijmegen
09.30.1.1 Coffee
09.45-12.30     Tom Heskes (SNN): 
              Confidence estimation for neural networks
12.30-13.30     Lunch
13.30-16.00 Computer practical 

Wednesday  4 April
Free for own exercise

Thursday 5 April: Amsterdam
09.30           Coffee
9.45-12.30      Ben Kröse: UvA
                Introduction to mixture modelling
                Nikos Vlassis:  UvA
                Greedy mixture learning
12.30-13.30     Lunch
13.30-16.00 Computer practical 

Friday 6 April: Amsterdam
09.30           Coffee
09.45-12.30     Nikos Vlassis: UvA: 
                Linear feature extraction
12.30-13.30 Lunch
13.30-15.00     Computer practical (Computer-assisted instruction)

REGISTRATION FORM

Name:............................Init.:.......Tit:..........

Institute:....................................................

Department:................................................

Address:....................................................

Postcode:...............Place:.............................

Telnr:....................Faxnr:............................

Email:.......................................................

Remarks:...................................................
..............................................................
..............................................................
Date:.......................................................


O       PhD-students (no charge)
O       Other university employee (dfl. 200,-)
Industry per day(s) (dfl. 500,- per day)
        O   Monday              O   Tuesday
        O   Wednesday   O   Friday
O       Industry  - complete course (dfl. 1.500,-)

Signature:

During the course, coffee and tea will be provided. The local cafeteria
will be available for lunch (not included).

For further information:         
Ben Kröse       
Email: krose@wins.uva.nl         
Phone: 020 525 7461     
http://www.science.uva.nl/~krose/asci_nn


Registration:
SNN, Foundation for Neural Networks 
KUN, MBFYS, 231
Annet Wanders
Geert Grooteplein 21
6525 EZ Nijmegen

Fax:    024 354 1435 
Phone 024.361 42 45
Email: snn@mbfys.kun.nl


--
Mailaddress:
Foundation for Neural Networks (SNN)    SNN,  - 231
http://www.snn.kun.nl                  Geert Grooteplein 21
Tel: +31 243614245                      6525 EZ Nijmegen
Fax: +31 243541435                      The Netherlands

0r:
SNN, 231
PO Box: 9101, 6500 HB Nijmegen