Copy Link
Add to Bookmark
Report
Neuron Digest Volume 12 Number 06
Neuron Digest Thursday, 7 Oct 1993 Volume 12 : Issue 6
Today's Topics:
Summer Institute
Re: Info on Neural Network based Multivariate Analysis
BP on Prediction of Protein Structure Problem
Protein Structure Prediction
Postdoctoral position
Model problem for feedforward networks
Deadline reminder: Music/arts special issue
Research Job - Switzerland
Research post - Treebank-translation
Stochastic Neural Networks for Optimization problems
Positions in BIOCOMPUTING
Send submissions, questions, address maintenance, and requests for old
issues to "neuron-request@psych.upenn.edu". The ftp archives are
available from psych.upenn.edu (130.91.68.31). Back issues requested by
mail will eventually be sent, but may take a while.
----------------------------------------------------------------------
Subject: Summer Institute
From: Terry Sejnowski <terry@helmholtz.sdsc.edu>
Date: Tue, 28 Sep 93 17:49:21 -0800
SUMMER INSTITUTE IN COGNITIVE NEUROSCIENCE
at The University of California, Davis
The 1994 Summer Institute will be held at the University of
California, Davis, from July 10 through 23. The two week course will
examine how information about the brain bears on issues in cognitive
science, and how approaches in cognitive science apply to neuroscience
research. A distinguished international faculty will lecture on current
topics in brain plasticity, strategies of neuroencoding, and evolution.
Laboratorites and demonstrations will provide practical experience with
cognitive neuropsychology experiments, connectionist/computational
modeling, and neuroimaging techniques. At every stage, the relationship
between cognitive processes and underlying neural circuits will be
explored. The Foundation is providing room/board and limited support for
travel.
Faculty Include:
Richard Andersen, Max Bear, Ira Black, Kenneth H. Britten, Simon Baron
Cohen, Leda Cosmides, Randy Gallistel, Michael S. Gazzaniga, Charles
Gilbert, Charles M. Gray, Eric Knudsen, Peter Marler, Michael Merzenich,
Reid Montague, Steven Pinker, V. Ramachandran, Gregg Recanzone, Barry
Richmond, Mitch Sutter, Timothy Tonini, John Tooby, and many others.
For information and applications please write to:
McDonnell Summer Institute in Cognitive Neuroscience
Center for Neuroscience, 1544 Newton Court
University of California, Davis
Davis, California 95616 USA
APPLICATIONS MUST BE RECEIVED BY JANUARY 15, 1994
------------------------------
Subject: Re: Info on Neural Network based Multivariate Analysis
From: URPANI D <dju@stan.xx.swin.OZ.AU>
Date: Thu, 30 Sep 93 16:18:37 -0100
Neuron Digesters
This is a plea for information. I am a PhD student at Swinburne Uni,
Melbourne, Australia. My interests are neural net applications in the
process industry with particular reference to aluminum smelting.
The idea is to find ways and means to make the process operate in a more
efficient envelope. Approx. 15 sensor measurements are currently logged
daily. My first step is to try to cluster there measurements into groups
so that later on I can assign a process conditon or state to each one.
Each condition can then be tagged with a process performance index such as
kg Al/kWH. I have been using hierarchical cluster analysis in the hope of
achieving this so I will have set of input vectors (15 dimensional) and a
fair,good, bad etc process condition attached to them ie I can then
progress form an unsupervised to a supervised learning situation.
However cluster analysis does not seem to be discriminating as well as I
hoped and I think I will have to do the clustering job using NN
(SOM's, LVQ's etc). I suspect there could be some problems associated
with the time series nature of the data (variables autocorrelated etc).
I wonder if anyone of you have come across a similar problem and if you
could volunteer some help.
Thank you.
Regards,
David Urpani// dju@stan.xx.swin.OZ.AU //
------------------------------
Subject: BP on Prediction of Protein Structure Problem
From: pb@cse.iitb.ernet.in (Pushpak Bhattacharya)
Date: Fri, 01 Oct 93 23:44:20 +0700
We are trying to train a BP neural net (PDP-3 package of Rumelhart and
Mclleland is being used) with the input as the primary structure and the
output as the conformational state of the amino acids. The net has 65
binary inputs and 3 binary outputs. The number of patterns to be trained
are more than eight thousand. We have 20 neurons in the hidden layer.
Now, just after one epoch the hidden neurons go into saturation state -
all their outputs are 1. At the same time the output neurons also go into
saturation. All their outputs become 0. We understand that the primary
reason for this is that A LARGE number of patterns are being trained on a
net with large no of inputs. The weights to the o/p layer keep on getting
the same kind of "Credit", so that their values go on decreasing to large
negative values.
Could anybody suggest a way out of this difficulty - which to our mind is
a result of applying BP on a large-size real life problem ? (those
applying BP on large problems will appreciate our difficulty !).
------------------------------
Subject: Protein Structure Prediction
From: pb@cse.iitb.ernet.in (Pushpak Bhattacharya)
Date: Fri, 01 Oct 93 23:49:24 +0700
Oops! we forgot to mention the email addr
pb@cse.iitb.ernet.in (Pushpak Bhattacharyya, CS & E , IIT Bombay)
susmita@cse.iitb.ernet.in (Susmita De, CS & E, IIT Bombay
Thanking you,
Pushpak Bhattacharyya
Susmita De
------------------------------
Subject: Postdoctoral position
From: axon@cortex.rutgers.edu (Ralph Siegel)
Date: Fri, 01 Oct 93 21:17:56 -0500
PLEASE POST
Postdoctoral Fellow. Analysis of visual structure-from-motion in
primates. These studies combine visual psychophysics and single unit
recording in awake behaving monkeys. Single unit recordings are being
made in visual association cortex using carefully controlled visual
stimuli. Laboratory also has ongoing human psychophysical studies with
normal and cortically compromised human subjects as well as computational
studies. Recent graduates who are changing fields from either cellular or
computational neuroscience to behavioral and physiological studies are
encouraged to apply. Computer expertise useful, but not necessary.
Superb experimental and computational facilities in a multi-disciplinary
research center. NY-NJ Metro area.
Contact: Ralph Siegel
Center for Molecular and Behavioral Neuroscience
Rutgers, The State University
197 University Avenue
Newark, NJ 07102
phone: 201-648-1080 x3261
fax: 201-648-1272
email: axon@cortex.rutgers.edu
Term: 24 months, beginning 10/1/93 or later
Salary: NIH level and supplement
Please send statement of research interests, curriculum vitae, and
names of three references.
------------------------------
Subject: Model problem for feedforward networks
From: maxim@wisdom.weizmann.ac.il (Kovalenko Maxim)
Date: Sat, 02 Oct 93 20:31:34 +0100
Hello,
I am developing a learning paradigm for feedforward networks
that presumably improves the convergence properties of
the currently available algorythms.
However, one of the drawbacks of the approach is that
input/output patterns should have the "space" nature,
i.e. have local correlation like in images or time series.
Could anyone point me to a good (practically significant)
model problem that involves this specific nature of
input/ouput data and is known to be solved slowly using
algorythms like back-propagation?
I would appreciate any related references or comments.
Thank you for your attention,
Maxim Kovalenko
_________________________________________________________________________
|
Maxim L. Kovalenko, | e_mail: maxim@wisdom.weizmann.ac.il
Department of Apllied Mathematics, | fax: 972-8-344122
The Weizmann Institute of Science, | tel.: 972-8-343668
Rehovot, 76100 |
Israel |
------------------------------
Subject: Deadline reminder: Music/arts special issue
From: "Peter M. Todd" <ptodd@spo.rowland.org>
Date: Sat, 02 Oct 93 18:05:02 -0500
**** PLEASE DISTRIBUTE ****
MUSIC AND CREATIVITY
Call for Papers for a Special Issue of Connection Science--
Reminder of approaching deadline
The October 15 deadline for submissions to the special issue of Connection
Science on network applications in music, arts, and creativity, is fast
approaching. We seek full-length papers on empirical or theoretical work in
the areas of modelling musical cognition; network composition, choreography,
or visual creation; integration of high- and low-level musical or artistic
knowledge; cross-modal integration (e.g. rhythm and tonality); developmental
models; cross-cultural models; psychoacoustic models; relationships between
music and language; and connections to cognitive neuroscience. We also
welcome shorter research notes up to 4000 words in length covering ongoing
research projects. For a complete call for papers and author guidelines, or
to submit a paper (five copies), contact the Special Issue Editors:
Niall Griffith
Department of Computer Science,
University of Exeter,
Prince of Wales Road,
Exeter,
EX4 4PT, England.
E-mail: ngr@dcs.exeter.ac.uk
Peter M. Todd
The Rowland Institute for Science
100 Edwin H. Land Boulevard
Cambridge, MA 02142 USA
E-mail: ptodd@spo.rowland.org
------------------------------
Subject: Research Job - Switzerland
From: Christian Lehmann <Christian.Lehmann@di.epfl.ch>
Date: Mon, 04 Oct 93 10:39:26 +0000
University of Lausanne:
Graduate student position (Doctorant) available in October 1993 at the
Institute of Physiology
Topic: Spatial and temporal processing in neural networks
The student will be integrated in an electrophysiology group working with
simultaneous single unit recordings. A tight collaboration with the Swiss
Federal School of Technology (EPFL) will provide the latest technical
facilities. It is expected that she/he will acquire in-depth knowledge in
the fast growing field of neural networks in order to develop and test
original ideas on information processing in the brain. A good background in
mathematics, physics, and biology as well as knowledge of at least one
higher programming language is recommended.
Our Ph.D. program extends over a duration of three years minimum. The
minimum salary ranges between US$19,000 and 24,000/year.
Please send applications (curriculum vitae and letters of recommendations
of two academic referees) to or get further information from:
Dr. Alessandro Villa or Dr. Yves de Ribaupierre, UNIL Institute of
Physiology, Rue du Bugnon 7, CH-1005 Lausanne, Switzerland.
Tel. ++41-21-313.2809
FAX ++41-21-313.2865
E-mail: villa@ulmed.unil.ch
------------------------------
Subject: Research post - Treebank-translation
From: E S Atwell <eric@scs.leeds.ac.uk>
Date: Mon, 04 Oct 93 16:10:05 +0000
The following is an advert for a Research Fellowship and one or more
Visiting Fellowships attached to a project to map between a number of
English Corpus annotation schemes. The exact method for mapping between
syntax trees of two given parsing schemes is open - presumably one way is
to `hand-annotate' a common sample (we call this a Multi-Treebank as each
sentence has more than one parse-tree) and then set a Neural Net to learn
the mapping between corresponding parses. Anyone who'd like to do this
is welcome to seek further details, and maybe even apply for the 3-year
post or just visit Leeds temporarily...
Eric Atwell
cf:
THE UNIVERSITY OF LEEDS
SCHOOL OF COMPUTER STUDIES - AI Division
Centre for Computer Analysis of Language And Speech (CCALAS)
RESEARCH FELLOW IN COMPUTATIONAL LINGUISTICS
The above SERC-funded post is available immediately for a fixed period of
three years to work on a project in natural language processing, involving
mapping between the syntactic annotation schemes of different ragged and
parsed corpora, including LOB, Brown, London-Lund, UPenn, SEC, ICE, British
National Corpus.
A PhD or equivalent expertise in Linguistics, Computer Science or Artificial
Intelligence is required; experience of corpus-based computational linguistics
and the syntactic models of one or more of these corpora is preferred.
Salary will be on the scale for Research Staff Grade IA (#12,828 - #20,442)
according to qualifications and relevant experience.
Informal enquiries about the post may be made to Eric Atwell, tel 0532 335761,
fax 0532 335468, email eric@scs.leeds.ac.uk or Clive Souter, tel 0532 335460,
email cs@scs.leeds.ac.uk
Application forms and further particulars may be obtained from the Personnel
Office (Academic Section), The University, Leeds LS2 9JT, England, tel 0532
335771 quoting reference no 48/105.
Closing date for applications: November 1st 1993.
The University of Leeds promotes an equal opportunities policy
****************************************************************************
SERC VISITING FELLOWSHIPS
SERC may also fund one or more Visiting Fellowships to support leading
researchers from other Institutions who can contribute towards the project,
visiting Leeds University for between a month and a year. We would
particularly welcome researchers with in-depth experience of one or more of
the tagging and/or parsing schemes, to advise us in the creation of the
detailed mapping algorithms, and the Multi-Tagged Corpus and MultiTreebank. If
you are interested in visiting CCALAS as a project advisor, please contact
Eric Atwell (eric@scs.leeds.ac.uk) and/or Clive Souter (cs@scs.leeds.ac.uk).
****************************************************************************
PROJECT SUMMARY: Mapping Between Corpus Annotation Schemes
Several alternative tagged and parsed Corpora of English exist, including LOB,
Brown, London-Lund, UPenn, SEC, ICE, British National Corpus, each with its
own tagset and/or parsing scheme. A tagged or parsed Corpus has many
applications, such as training linguistic constraint models for improved
speech recognition; however users cannot combine Corpus training sets into a
single language model, as the annotation schemes are incompatible.
This project will design a set of tag- and tree-transducers or algorithms for
mapping between the main corpus annotation schemes. This will allow users of
one Corpus to view other Corpora as enlargements of their training set. One
tagset and parsing scheme will be our 'base' or interlingua, and transducers
will be built between this interlingua and the other annotation schemes. A
relatively small test corpus will be annotated with all the schemes under
consideration; we will investigate the use of the resulting Multi-tagged
Corpus and Multitreebank as a standard evaluation benchmark for taggers and
parsers.
****************************************************************************
COMPUTATIONAL LINGUISTICS RESEARCH AT LEEDS UNIVERSITY
The School of Computer Studies at Leeds provides excellent broad background
support for research; we were graded 4(A) by UFC/HEFCs, and NLP makes an
important and growing contribution to the School's research profile. At Leeds
the Centre for Computer Analysis of Language And Speech (CCALAS), with Eric
Atwell as Director and Peter Roach and Clive Souter as Deputy Directors,
provides a focus for a broad range of corpus- and dictionary-based research
including word-sense semantic disambiguation and tagging (Demetriou, Jost,
Atwell), grammar-based reasoning (Mott, Silver), speech act theory (Holdcroft,
Wallis, Wynne, Millican), probabilistic parsing (Pocock, O'Donoghue, Atwell,
Souter, Hogg), corpus collocation analysis (Howarth, Cowie, Davidson), corpus
annotation (Atwell, Roach, Souter, Arnfield, Ghali, Bull), grammatical
inference and clustering (Hughes, Tarver, Atwell) speech recognition (Roach,
Ueberla, Kirby, Moore, Lockhart, Mair, Sergant), speech synthesis (Scully,
Roach), handwriting recognition (Hanlon, Boyle, Bushofa), text generation
(Cole, Grierson, Tawalbeh), human-computer interaction (Crow), computers in
language teaching and linguistics (Davidson, Fox, Roach, Hunter, Shivtiel),
computers in lexicography (Roach, Setter, Cowie, Atwell, Souter).
****************************************************************************
Leeds University has over 15,000 students and 2,000 academic and research
staff, making it one of the largest in Britain. Leeds is half-way between
London and Edinburgh, linked by rail, motorway and air to the rest of the UK
and Europe. It is the 20th largest city in the European Community, with the
excellent arts, sport and other social facilites expected of a growing,
multi-cultural metropolis; but it is also close to four National Parks. More
background information on the Project, CCALAS, the University, and Leeds and
its environs can be found in the Further Particulars from the Personnel
Office.
------------------------------
Subject: Stochastic Neural Networks for Optimization problems
From: suchi@pollux.cs.uga.edu (Suchi Bhandarkar)
Date: Mon, 04 Oct 93 16:08:49 -0500
Could someone give me references on Stochastic Neural Network-based
approaches to combinatorial and continuous variable optimization? Please
e-mail your responses to "suchi@cs.uga.edu" Thank you very much.
Sincerely,
Suchi Bhandarkar
University of Georgia
E-mail: suchi@cs.uga.edu
------------------------------
Subject: Positions in BIOCOMPUTING
From: Soren Brunak <brunak@cbs.dth.dk>
Date: Tue, 05 Oct 93 12:04:07 +0000
Positions in BIOCOMPUTING at the Danish Center for Biological Sequence
Analysis, Department of Physical Chemistry, The Technical University of
Denmark (Lyngby).
A number of pre- and post-doctoral positions are available at the newly
formed Center for Biological Sequence Analysis. They have a duration of
one, two and three years, starting late 1993 or early 1994. The center
is funded by a five-year grant from the Danish National Research
Foundation and conducts an active research program in biomolecular
sequence and structure analysis with emphasis on novel adaptive
computational strategies. The Technical University of Denmark is
situated in Lyngby just outside Copenhagen.
The center offers employment to researchers with a background primarily
in the natural sciences, molecular biology, genetics, chemistry and
physics. We seek individuals with additional competence and interest in
areas of computer science, but not with this area as the main subject
of expertise. Priority will be given to younger scientists with
experience in some of the following areas (in alphabetical order):
experimental molecular biology, information theory and statistics,
mathematical analysis, neural computation, protein folding, physics of
computation and complex systems, and sequence analysis.
In a wide range of projects the center collaborates with national and
foreign groups using novel adaptive computational methods many of which
have received attention in the biocomputing context only recently. The
center is characterized by the use of new approaches both regarding the
algorithmic aspect of the simulation methods as well as the use of
advanced hardware. A wide range of parallel and cluster computing
environments is available locally at the center; a nearby supercomputer
center offers easy access to CRAY and Connection Machine facilities.
Among the research topics are pre--mRNA splicing, recognition of
vertebrate promoters, RNA folding, protein structure prediction,
proteolytic processing of polyproteins, signal peptide recognition,
phylogenies, global multiple sequence alignment and dedicated sequence
analysis hardware. The results are evaluated through intensive exchange
with experimentalists.
For further information, feel free to contact us at the address below.
Applicants should send their resumes to
Soren Brunak
Center director
Center for Biological Sequence Analysis
Department of Physical Chemistry
The Technical University of Denmark
Building 206
DK-2800 Lyngby
Denmark
Tel: +45-42882222, ext. 2477
Fax: +45-45934808
Email: brunak@cbs.dth.dk
------------------------------
End of Neuron Digest [Volume 12 Issue 6]
****************************************
