Copy Link
Add to Bookmark
Report
Neuron Digest Volume 02 Number 19
NEURON Digest Thu Aug 13 18:11:01 CDT 1987 Volume 2 / Issue 19
Today's Topics:
financial applications
ONR Program Announcement for Neural Networks Research
Another Parallel Machine
Computational Neuroscience Symposium
Conference - Fifth International Machine Learning Conference
Symposium Announcement - 3rd NBCA
----------------------------------------------------------------------
Date: Fri 24 Jul 87 17:08:32-EDT
From: "John C. Akbari" <AKBARI@cs.columbia.edu>
Subject: financial applications
anyone out there in net land considering applications of neurocomputing
techniques to pattern detection and optimization in trading or other
financial applications? if so, send mail.
John C. Akbari
JCA and Associates
380 Riverside Drive, No. 7D
New York, New York 10025 USA
Tele. 212.662.2476
ARPANET & Internet akbari@CS.COLUMBIA.EDU
BITnet akbari%CS.COLUMBIA.EDU@WISCVM.WISC.EDU
UUCP columbia!cs.columbia.edu!akbari
-------
------------------------------
Date: 24 Jul 1987 17:48-EDT
Subject: ONR Program Announcement for Neural Networks Research
From: CLAU@a.isi.edu
OFFICE OF NAVAL RESEARCH (ONR) PROGRAM ANNOUNCEMENT
ELECTRONIC NEURAL SYSTEMS and BIOLOGICAL NEURAL NETWORKS
The Office of Naval Research (ONR) announces two new Accelerated Research
Initiatives to start supporting basic research in FY89 in electronic and
biological neural networks.
The objective of the Electronic Neural Systems program is to study/implement
highly efficient neural based electronic architectures utilizing new analog
VLSI technology by exploiting the latest understanding of biological neural
networks, and to develop a deeper mathematical understanding of the nonlinear
behavior of the fundamental processes in biological neural networks. In a
companion program, the objective of the Biological Neural Networks program is
to provide information necessary to permit new biologically derived
computational approaches to currently difficult or intractable problems of
Navy relevance, and to stimulate new interdisciplinary approaches to study
biological information processing systems (neural networks).
Research proposals are now accepted by ONR. All proposals received before
June 1, 1988 will be considered for funding beginning in October 1988. For
more information, please contact:
ELECTRONIC NEURAL SYSTEMS
Dr. Clifford Lau (818)795-5971 Electronics CLAU@A.ISI.EDU
Dr. Lyle Broemeling (202)696-4314 Mathematics
BIOLOGICAL NEURAL NETWORKS
Dr. Joel Davis (202)696-4744 Biology
Dr. Alan Meyrowitz (202)696-4302 Computer Science
Office of Naval Research
800 N. Quincy Street
Arlington, VA 22217-5000
------------------------------
Date: 9-AUG-1987 10:37
From: STROM@OREGON-GRAD.CSNET
Subject: Another Parallel Machine
The OGC iPSC-based Neural Network Simulation System
Casey Bahr and Dan Hammerstrom
Oregon Graduate Center
The Department of Computer Science/Engineering at the Oregon
Graduate Center has developed an extensive system for the simula-
tion of arbitrary neural network structures on the Intel iPSC
hypercube. This system consists of a compiler that takes as
input a Network Description Language, NDL (NooDLe), and generates
a well-defined, low-level generic network specification, Beaver-
ton Intermediate Form, BIF; a network partition utility (network
node to iPSC processor mapper); and an iPSC based neural network
emulator, ANNE (Another Neural Network Emulator).
The user describes the structure of the network very quickly
in NDL, which is based on Scheme (a lexically scoped, typed ver-
sion of Lisp). The user then deals with the components of the
network as objects and combines them accordingly. Arbitrary,
stochastic interconnectivity is possible. The functions of the
particular nodes (and sites at the nodes) for computation and
learning are specified by pointers to C procedures. Internal to
these C functions the user code accesses network parameters via
standardized data structures and calls to ANNE. OGC will eventu-
ally provide a variety of standard functions for typical network
architectures (back-propagation, adaptive-resonance, etc.), but
the user can easily add his own.
The NDL source is compiled into BIF and run through the
mapper, which, using a PAD (Physical Architecture Descriptor)
file that describes the user's machine, partitions the network
among the iPSC processors. This mapped BIF is then read by ANNE.
ANNE's purpose is to act as a testbed and debugger for the
variety of neural network models describable by BIF. ANNE is not
intended to model any particular architecture design on which
these networks might be mapped, but rather it is designed to run
neural nets in an expedient manner in order to examine their
operating characteristics.
What ANNE provides the designer of experimental neural net-
works is a virtual machine over a distributed memory multiproces-
sor computer that is specifically designed for the characteris-
tics of neural network models. As part of the complete design
environment it supplies a scanner/builder to construct a neural
network from BIF, a message passing mechanism for coordinating
communication between network nodes on different iPSC processors,
and a special timing and synchronization system (which is user
controlled). In addition, the user has the ability to examine,
modify, or save pertinent data within the network, including the
entire BIF description (plus network state) of the network at any
point in the simulation.
ANNE reads the mapped BIF specification to construct the
network under consideration. Included with the BIF specification
are the C routines that specify the behavior of the network ele-
ments. Each of network node may have its own user-defined func-
tion. The ANNE node image, including these functions, is loaded
into each iPSC processor.
Although most of the code is written, ANNE is not yet fully
operational. We expect to begin simulating networks within the
next 2-3 months, so we cannot as yet provide performance figures.
This software was developed under a joint grant from the Office
of Naval Research and the Air Force Office of Scientific
Research, and we will gladly give it to other research groups
(caveat emptor). ANNE was written to be easily extensible to
future, more powerful versions of the Intel system. The main
goal of the Cognitive Architecture Project at OGC is to develop
special silicon architectures for neural/connectionist network
emulation. ANNE was developed as a tool for this research, as a
result, we have made many design decisions based on this intended
use (such as using an intermediate form).
------------------------------
Date: Thu, 30 Jul 87 23:59:32 EDT
From: Terry Sejnowski (JHU|mike) <terry@brl.arpa>
Subject: Computational Neuroscience Symposium
Wednesday, September 2, 1987
PUBLIC SYMPOSIUM ON COMPUTATIONAL NEUROSCIENCE
Whitman Auditorium
Marine Biological Laboratory
Woods Hole, Massachusetts
1:30 PM Terrence Sejnowski, Johns Hopkins University
Introduction
2:00 PM Leif Finkel, Rockefeller University
Functional Plasticity in the Topographical
Organization of Cortical Maps
3:00 PM Richard Andersen, Massachusetts Institute of Technology
Spatial Representations in Parietal Cortex:
Models and Experiments
4:00 PM Rodolfo Llinas, New York University
Sensory-Motor Processing in the Cerebellum:
Responses from Neural Populations
------------------------------
Date: Fri, 10 Jul 87 11:01:21 EDT
From: laird@caen.engin.umich.edu (John Laird)
Subject: Conference - Fifth International Machine Learning Conference
[This is an old message from AIList Digest - MTG]
CALL FOR PAPERS
FIFTH INTERNATIONAL CONFERENCE ON MACHINE LEARNING
Ann Arbor, Michigan
June 12-15, 1988
The Fifth International Conference on Machine Learning will be held at the
University of Michigan, Ann Arbor, during June 12-15, 1988. The goal of the
conference is to bring together researchers from all areas of machine learning.
The conference will have open attendance and registration fees.
REVIEW CRITERIA
In order to ensure high quality papers, each submission will be reviewed
by two members of the program committee and judged on clarity, significance,
and originality. The best papers will be published in the proceedings, and
their authors will be invited to give a talk on their work or describe it at
a poster session. All submissions should contain new work, new results, or
major extensions to prior work. Summaries and overviews are discouraged.
The ideal paper will present a clear description of the learning task being
addressed and the proposed solution to that problem. If the paper describes
a running system, it should explain that system's representation of inputs
and outputs, its performance component, and its learning methods. It should
include a detailed example, as well as relate the work to earlier research.
Most important, all papers should include some evaluation of the work in the
form of substantive results. Papers are not required to take this form, but
authors are strongly encouraged to follow this format.
SUBMISSION OF PAPERS
Each paper must have a cover sheet with the title, authors' names, primary
author's address and telephone number, and an abstract of about 200 words. The
cover page should also give three keywords that specify the problem area,
general approach, and evaluation criteria. Some examples of each are:
PROBLEM AREA GENERAL APPROACH EVALUATION CRITERIA
Concept learning Genetic algorithms Empirical evaluation
Learning and planning Empirical methods Theoretical analysis
Language learning Explanation-based Psychological validity
Learning and design Connectionist
Machine discovery Analogical reasoning
The body of the paper must not exceed 13 double-spaced pages in twelve point
font, including figures but excluding references. Authors should send four
copies of their papers to:
Machine Learning Conference
Cognitive Science and Machine Intelligence Laboratory
The University of Michigan
904 Monroe Street
Ann Arbor, MI 48109-1234
Internet: ml88@csmil.umich.edu
The deadline for submission of papers is January 15, 1988. Authors will be
notified of acceptance by March 1, 1988. Final camera-ready copies of the
papers will be due April 1, 1988.
Organizing Committee
J. E. Laird (chairman) University of Michigan
J. H. Holland, S. L. Lytinen, G. M. Olson University of Michigan
J. G. Carbonell, T. M. Mitchell Carnegie-Mellon University
P. Langley University of California, Irvine
R. S. Michalski University of Illinois
------------------------------
Date: Fri, 24 Jul 87 11:08:44 CDT
From: gately@resble.ti (Michale T. Gately)
Subject: Symposium Announcement - 3rd NBCAOB
What : 3rd Annual Symposium on Networks in Brain and Computer Architecture
When : September 18-20, 1987
Where: North Texas State University, Denton, Texas
University Union Lyceum
Sponsored by: The NTSU Center for Network Neuroscience
Supported by: The Communities Foundation of Texas,
Rockwell International Corporation, and
North Texas State University
Chairman: Guenter Gross, Center for Network Neuroscience, NTSU
Program:
FRIDAY, September 18
0800-0900 Registration and Continental Breakfast
0900-0910 Welcome and opening remarks: Rollie Schafer
(Associate Dean for Science and Technology - NTSU)
0910-0940 Keynote Address and Plenary Lecture: open
SESSION 1: Network Analysis in Neurobiological Systems
0950-1030 Tritonia Swimming - A Model System for Integration
within Rhythmic Motor Networks: Peter Getting
(University of Iowa)
1030-1050 Coffee Break
1050-1130 Parallel Processing and Self Organization in
Invertebrate Nervous Systems: George Mpitsos (Oregon
State University)
1140-1220 Network Function in the Olfactory Cortex: James
Bower (California Institute of Technology)
1230-1330 Lunch
1400-1440 Cognitive Processes in Context of Basal Ganglia
Function: Donald J. Woodward (University of Texas
Health Science Center at Dallas)
1450-1530 Gating of Sensory Processes: John K. Chapin (Hahneman
Medical School, Philadelphia)
1540-1600 Coffee Break
1600-1640 The Role of the Hippocampus in Memory: Samuel Deadwyler
(Bowman Gary University)
1650-1730 Network Feature Extraction via Mulitelectrode Burst
Pattern Analysis: Guenter W. Gross (Center for Network
Neuroscience, North Texas State Universty)
1830-1930 OPEN BAR - Entertainment by the NTSU jazz combo
1930-2030 DINNER - Entertainment by harpist Ellen Ritscher
Special Lecture - Is Network Modeling Applicable to
Human Behavior?: Elliot Ross M.D. (Department of
Neurology, University of Texas Health Science Center
at Dallas)
SATURDAY, September 19
0800-0900 Breakfast Buffet
0900-0940 Electrical Synapses in Mammalian Hypothalmic Nuclei:
Implications for Network Properties and Analysis:
Glenn I. Hatton (Michigan State University)
0950-1030 The Mathematics of Neurons: Frank C. Hoppensteadt
(Michigan STate University)
1040-1100 Coffee Break
SESSION 2: Application and Implementation of Neural Network Models
1100-1140 Olfactory Bulb Network Theory: Walter Freeman
(University of California, Berkeley)
1200-1300 Lunch
1330-1410 Neural Network Application for Acoustic Signal
Recognition: Richard Burne (Bendix Aerospace Technical
Center)
1420-1500 Applications of Neural Network Computing: Harold Szu
(Naval Research Laboratory)
1550-1610 Coffee Break
1610-1650 Application of Neural Netwroks to Recognition of
Handwritten Numerals: F. Kamangar (University of Texas
at Arlington)
1700-1740 New Learning Mechanisms for the Boltzmann Machine:
Carsten Petersen (M.C.C., Austin)
1800-2000 Cocktail Hour and Buffet, Entertainment: TWU Choraliers
1900-2100 Poster Session and Informal Discussion
Sunday, 20 September 1987
0800-0900 Breakfast Buffet
SESSION 3: Workshop: Multichannel Recording and Display Technology
0900-1200 Workshop: Multichannel Recording and Display Technology
Panalists
Spence BeMent (University of Michigan)
John Chaplin (Hahneman Medical School, Philadelphia)
George Gerstein (University of Pennsylvania)
Chris Myer (University of Texas Health Science Center at Dallas)
David Raymond (Center for Network Neuroscience, NTSU)
Allen Selverston (University of California, San Diego)
Bruce Wheeler (University of Illinois)
John White (Center for Network Neuroscience, NTSU)
0900-1200 Poster Session
1200-1300 Buffet Lunch
Registration fee: $55 professional
$5 student
Send Registration fee and information (name address and telephone
number) to the address below. Please make checks payable to
NTSU-Networks.
Call for Posters - If you are interested in setting up a 4'x5.5'
poster, please sent a note indicating the Title, Author(s),
Affiliation, Address, and Telephone number to the address listed
below (do not send the poster).
Lodging arrangements can be made through the address/phone number
listed below. Prices range from 14.75 per night (Single campus
housing) to $54/night (Double room at Denton Sheraton).
All meals indicated in the program are available to attendees.
The total cost of all meals is $45. Each meal can be purchased
separately.
Further information:
Networks Conference Management
P.O. Box 5344
North Texas State University
Denton TX 76203
(817) 565-3481
------------------------------
End of NEURON-Digest
********************
