Copy Link
Add to Bookmark
Report
Neuron Digest Volume 13 Number 23
Neuron Digest Thursday, 7 Apr 1994 Volume 13 : Issue 23
Today's Topics:
ICNN '94 Call For Participation
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) in pub/Neuron-Digest or by
sending a message to "archive-server@psych.upenn.edu".
----------------------------------------------------------------------
Subject: ICNN '94 Call For Participation
From: Dennis W. Ruck <druck@afit.af.mil>
Date: Mon, 14 Mar 1994 08:31:02 -0500
--------------------------
REGISTRATION INFORMATION
--------------------------
IEEE WORLD CONGRESS ON COMPUTATIONAL INTELLIGENCE
Orlando, FLA
June 26-July 2, 1994
IEEE International Conference on Neural Networks
Third IEEE International Conference on Fuzzy Systems
The IEEE Conference on Evolutionary Computation
Special Plenary Symposium
"Computational Intelligence: Imitating Life"
-> Over 1600 Refereed and Invited Presentations <-
-> 43 Cutting Edge Plenary Presentations <-
- -> Eighteen Cutting Edge Tutorials on Newest Innovations <-
-> Current Technology Exhibits <-
*** * ** ** *** ** **** ****
**** * * ** ***************** *
***** * * ** ************ *
* X ************** *
**** ****** ** ** ******
***** ******* * ***
** *** * **
* *** ***
* * ***** *
* * *
Sponsored by the IEEE Neural Networks Council
Exhibits organized by SPIE
********************************************
FREE CD-ROMS OF ICNN'93 AND FUZZ-IEEE
FOR THE FIRST 1000 REGISTRANTS
********************************************
For additional information, please contact:
WCCI'94 Conference Office
Meeting Management
2603 Main Street, Suite 690
Irvine, CA 92714
Tel. (714) 752-8205
Fax (714) 752-7444
e-mail: 74710.2266@compuserve.com
CONTENTS OF THIS POSTING:
1. General Conference Information
2. Conference Registration Form
3. Tutorial Registration
4. Methods of Payment
5. Hotel Registration Form
6. Spouse Activities
7. Tutorial Titles
8. Tutorial Abstracts
******************************************
1. General Conference Information
******************************************
The 1994 IEEE World Congress on Computational Intelligence
consists of three IEEE International Conferences: The Third
IEEE International Conference on Fuzzy Systems, IEEE
International Conference on Neural Networks, and The IEEE
Conference on Evolutionary Computation. Over 1600 refereed
and invited papers will be presented in these Conferences as
well as to a special five day Symposium entitled
"Computational Intelligence: Imitating Life." This
Symposium will be held Monday, June 27 through Friday, July
1, 10:20AM to 12:20PM. The Congress Inaugural will be held
Tuesday, June 28, 6:30PM to 7:15PM.
Special Plenary Symposium
COMPUTATIONAL INTELLIGENCE: IMITATING LIFE
June 27 - July 1, 1994
For the first time in one meeting, the main threads of the
topics in computational intelligence are woven into a single
cohesive fabric. The Symposium addresses the exciting
emerging technologies and issues relating to biologically,
psychologically and linguistically motivated models that
exhibit various facets of computational intelligence.
Machine learning from data, neuro-fuzzy information
processing, approximate reasoning, vision/qualitory modeling
and evolutionary computation, are examples of computational
intelligence approaches addressed by Symposium speakers.
The Symposium provides a unique forum for cross-
fertilization between the areas of neural networks, fuzzy
logic, and evolutionary computation.
The Symposium consists of three public lectures, 10 plenary
talks and 30 mini-symposia presentations; covering neural
networks (21), fuzzy logic (13), and evolutionary
computation (9). Contributions include research that has
implications for further progress in the field, state-of-
the-art reviews, followed by discussions of important
applications in fields such as robotics and control, image
processing, vision, and biology. The presentations will be
highly focused but still tutorial.
Symposium speakers represent the top internationl
researchers and practioners of this cutting edge technology.
"Computational Intelligence: Imitating Life" companion
volume to this Symposium will be complimentary to each
registered participant of the Congress.
For more information, contact
Symposium Chair: Dr. Jacek Zurada
University of Louisville
Louisville, KY 40292, USA
PHONE: (502)852-6314, FAX: (502)852-6807
EMAIL: jmzura02@ulkyvx.louisville.edu
******************************************
2. CONFERENCE REGISTRATION FORM
******************************************
Indicate Conference Selection (you may attend sessions
from all three conferences)
( ) FUZZY ( ) NEURAL NETWORKS ( ) EVOLUTIONARY COMPUTATION
NAME
Last Name ________________________________
First Name ________________________________
Middle Initial ____________________________
_____________________ IEEE Membership Number
(needed to qualify for IEEE Member discount)
MAILING ADDRESS
City _________________________________________
State and ZIP (USA only)______________________
Country ______________________________________
Phone ________________________________________
FAX __________________________________________
e-mail _______________________________________
Information to appear on Badge:
Title (Circle One)
Ms Dr Prof no title other____________________
Full Name ______________________________________
Affiliation______________________________________
City/State/Country ______________________________
CONFERENCE REGISTRATION FEES:
before April 15, 1994 after April 15, 1994
IEEE Members $350.00 $425.00
Non-Members $420.00 $495.00
Students* $ 90.00 $150.00
* A letter from the Department Head to verify full-time
student status at the time of registration is required. At
the conference, all students must present a current student
ID with picture. Student registration does not include
social functions.
Full Conference Registration permits attendance at Congress
functions, the Symposium, technical sessions of all three
conferences, and the individual reception of the conference
selected.
Your registration fee includes the Proceedings (for the
conference selected: FUZZ-IEEE'94, ICNN '94 or ICEC '94) and
the Symposium Proceedings. A complete set of Proceedings
for all three conferences is available for an additional
$105. *Please note that Proceedings will be available after
the conference from IEEE, although the price for each will
increase at that time.
3. TUTORIAL REGISTRATION
Before April 15, 1994 After April 15,
1994
Regular Student Regular
Student
One Tutorial $225 $125 $300 $150
Two Tutorials $350 $200 $450 $225
Three Tutorials $475 $275 $600 $300
Four Tutorials $650 $350 $750 $375
Each additional $125 $ 75 $150 $75
Tutorial Selection (Circle desired tutorials)
1A 1B 2 3A 3B 4A 4B 5A 5B 6 7A 7B
8 9A 9B 10 11 12 13 14 15 16 17 18A 18B
Alternate Tutorial(s)
1A 1B 2 3A 3B 4A 4B 5A 5B 6 7A 7B
8 9A 9B 10 11 12 13 14 15 16 17 18A 18B
Payment(s):
Registration Fees U.S. $____________
Tutorial Fees U.S. $____________
All Three Proceedings ($105) U.S. $____________
Grand Total U.S. $____________
******************************************
4. Methods of Payment:
******************************************
$ CHECK. All check payments made outside of the USA must
be made on a USA bank in US dollars. Please make
check payable to WCCI '94
$ CREDIT CARDS. Only VISA, MC and Amex accepted. Payment
may be through e-mail. Registrations submitted
by fax or surface mail must include an authorized
signature.
( ) Visa ( ) M/C ( ) Amex
Name on Credit Card ______________________________________
Credit Card Number _______________________________________
Exp. Date ________________________________________________
Authorized Signature _____________________________________
All Conference (other than hotel) Registration material is
to be sent to
WCCI '94 Conference Office
Meeting Management
2603 Main Street, Suite 690
Irvine, CA 92714
USA
Tel. (714) 752-8205
Fax (714) 752-7444
e-mail: 74710.2266@COMPUSERVE.COM
5. HOTEL RESERVATION FORM
Reservation Payment may be made by Check or Credit Card.
Mail this form and payments to:
Walt Disney Sheraton World Dolphin
Attention: Reservations Department
1500 EPCOT Resort Blvd.
Lake Buena Vista, FL 32830
Please make checks payable to: Walt Disney World Dolphin
Check Desired Accommodations:
Single $145 _____
Double $145 _____
Non-Smoking _____
If requested bedding is not available, alternate bedding
will be assigned
NOTE: The standard rates during this time period start at
$255.00 per night, the conference rates offer a substantial
savings.
Arrival Date: __________ Departure Date: __________
Check-in Time: 3:00 pm Check-out Time 11:00 am
Name ______________________________________________________
Mailing Address ____________________________________________
City/State/Country/ZIP _____________________________________
___________________________________________________
___________________________________________________
Phone ___________________________________
FAX ___________________________________
Sharing Room With (if applicable) __________________________
Sheraton Club International Number (if
applicable)__________________________
Deadline Date: May 27, 1994
Please reserve before May 27, 1994, after this date, rooms
are subject to availability.
Please circle credit card type:
Visa M/C Amex DC CB ER DI JCB
Credit Card Number _____________________________________
Exp. Date ______________________________________________
Name on Credit Card ____________________________________
Authorized Signature ___________________________________
All reservations require a one night's deposit. Failure to
cancel your reservation 5 days prior to arrival will result
in forfeit of deposit.
Group rates can only be confirmed by using this reservation
form or calling the hotel directly. For any questions or
further information regarding your request , please call our
Reservations Office (toll free from the U.S.) at (800) 227-
1500, Fax # (407) 934-4710, or contact the hotel directly at
(407) 934-4000.
To avoid duplication, please do not mail this form if you
make your reservation by telephone or telefax.
******************************************
6. Spouse Activities
******************************************
* Tee times on three nearby WALT DISNEY WORLD Championshiop
Golf Courses.
* The Walt Disney World Dolphin connects by waterways and
walkways to EPCOT Center and the Disney-MGM Studios Theme
Park. Convenient complimentary Disney-operated
transportation ties directly to the MAGIC KINGDOM Park,
Pleasure Island, Typhoon Lagoon, the Disney Village
Marketplace, 3 nearby championship golf courses and other
areas of the Vacation Kingdom. The Dolphin offers the
following hotel amenities.
* Two acre swimming area with three pools, including a
themed grotto area with slide, waterfalls and whirlpool
area, lap pool plus lakeside white sand beach with special
activities and watercraft rental.
* Camp Dolphin, offering a wide range of youth activities
* Eight night-lit, hard tennis courts
* One-on-One personal fitness training under the guidance of
"Body by Jake", with sauna, whirlpool, weight room, and
exercise equipment
******************************************
7. TUTORIAL TITLES
******************************************
For the first time, the World Congress joins three IEEE
conferences on neural networks, fuzzy systems, and
evolutionary computation in a single comprehensive forum.
The forum presents two days of tutorials designed to provide
information and help attendees keep pace with developments
in paradigms that are guiding the development of models for
computational intelligence. WCCI tutorials will be held on
Sunday, June 26, 1994 and Wednesday, June 29, 1994. The
WCCI Organizing Committee reserves the right to cancel
tutorials and refund payment should registration not meet
the minimum number of persons per course.
SUNDAY JUNE 26, 1994
#1A Evolution Strategies: A Thorough Introduction
Professor Thomas Beack
#2 Genetic Algorithms and Their Applications
Dr. Lawrence "David" Davis
#3A An Introduction to Evolutionary Computation
Dr. David B. Fogel
#4A Genetic Programming
Dr. John R. Koza
#5A Genetics-Based Machine Learning in Rule-Based and
Neural Systems
Professor Robert E. Smith
#7A An Introduction to Fuzzy Logic
Professor James Bezdek
#9A Fuzzy Logic Applications to Artificial Intelligence and
Intelligent Control Systems
Dr. Enrique H. Ruspini
#10 Fuzzy Logic in Computer Vision
Professor James M. Keller
#11 Fuzzy Neurocomputations
Professor Witold Pedrycz
#12 Fuzzy Data Analysis
Professor Dr. Dr.h.c. Hans-Jurgen Zimmermann
#18A Learning Algorithms In Neural Networks
Professor Jacek M. Zurada
WEDNESDAY JUNE 29, 1994
#1B Evolution Strategies: A Thorough Introduction
Professor Thomas Beack
#3B An Introduction to Evolutionary Computation
Dr. David B. Fogel
#4B Genetic Programming
Dr. John R. Koza
#5B Genetics-Based Machine Learning
in Rule-Based and Neural Systems
Professor Robert E. Smith
#6 Genetic Algorithms:
Theoretical Foundations and Experimental Evaluation
Professor Darrell Whitley
#7B An Introduction to Fuzzy Logic
Professor James Bezdek
#8 Fuzzy Sets in Constraint Satisfaction
Dr. Didier Dubois
#9B Fuzzy Logic Applications to Artificial Intelligence
and Intelligent Control Systems
Dr. Enrique Ruspini
#13 Applications of Neural Networks to Virtual Reality
Professor Thomas P. Caudell
#14 Hybrid Systems: Neural, Symbolic, and Fuzzy
Professor Lawrence O. Hall and Professor Abraham Kandel
#15 Basics of Building Market Timing Systems:
Making Money with Neural Networks
Casimir C. Klimasauskas
#16 Practical Applications of Neural Network Theory
Dr. Robert Hecht-Nielsen
#17 Computational Studies of Biological Neural Networks:
Introduction and Applications
to Vision and Sensory-Motor Control
Professor Paolo Gaudiano
#18B Learning Algorithms in Neural Networks
Professor Jacek M. Zurada
******************************************
8. TUTORIAL ABSTRACTS
******************************************
#1 Evolution Strategies: A Thorough Introduction
Professor Thomas Beack
Computer Science Department, LS XI
University of Dortmund, Dortmund, Germany
In addition to Genetic Algorithms and Evolutionary
Programming, the Evolution Strategy (Evolutionsstrategie)
by Rechenberg and Schwefel forms the third major
representative of Evolutionary Algorithms. Since its
development in the 1960's at the Technical University of
Berlin (Germany) for solving experimental optimization
problems, the computer algorithm has been successfully
applied to numerous hard continuous parameter optimization
problems (an application field where Evolution Strategies
reveal their strengths in comparison to the more familiar
Genetic Algorithms).
The tutorial presents a thorough introduction to
Evolution Strategies, with special emphasis on the
following topics: history of evolution strategies,
detailed presentation and explanation of the algorithm,
genetic operators and parameter settings, self-adaptation
of strategy parameters, theory of evolution strategies,
selected application examples of evolution strategies,
evolution strategies for neural networks and fuzzy logic,
guidelines for practitioners, and comparison to genetic
algorithms and evolutionary programming.
#2 Genetic Algorithms and Their Applications
Dr. Lawrence "David" Davis Tica Associates
Cambridge, MA
Genetic algorithms are techniques for optimization and
machine learning that have been applied to a wide range of
real-world problems. This tutorial consists of an
overview of genetic algorithms, a discussion of techniques
for applying them, a survey of areas in which they have
been applied, and several application case studies.
Particularly stressed in the tutorial will be traditional
and nontraditional genetic algorithms for numerical
function optimization; the use of order-based genetic
algorithms for combinatorial optimization; and techniques
for hybridizing genetic algorithms with other optimization
algorithms.
#3 An Introduction to Evolutionary Computation
Dr. David B. Fogel
Natural Selection, Inc.
La Jolla, CA
The impact of evolutionary thinking on biology cannot be
underestimated. Indeed, many biologists have remarked
that the study of life cannot be conducted reasonably in
the absence of an evolutionary paradigm. But evolutionary
thought extends beyond an ordering principle of biology.
Evolution is a process that can be simulated on a computer
and used for solving difficult engineering problems and
gaining insight into natural evolved systems. This
tutorial, aimed at researchers in neural networks and
fuzzy systems, and beginners in the field of evolutionary
computation, will introduce methods of evolutionary
computation. These include genetic algorithms, evolution
strategies and evolutionary programming, as well as
related techniques. The fundamental philosophical
foundations of the methods will be discussed and
applications will be described, including synergistic
efforts of combining evolutionary optimization with
connectionist and fuzzy systems.
#4 Genetic Programming
Dr. John R. Koza
Consulting Professor
Computer Science Department,
Stanford University, Palo Alto CA
Genetic programming extends the genetic algorithm to
the domain of computer programs and genetically breeds
populations of computer programs to solve problems.
Genetic programming can solve problems of system
identification, optimal control, pattern recognition,
equation solving, game playing, optimization, and
planning. Starting with hundreds or thousands of randomly
created programs, the population is progressively improved
by applying Darwinian fitness proportionate reproduction
and crossover (sexual recombination).
Many problem environments have regularities,
symmetries, and homogeneities that can be exploited in
solving the problem. The recently developed facility of
automatic function definition enables genetic programming
to dynamically decompose a problem into simpler
subproblems, solve the subproblems, and assemble original
problem. Experimental evidence suggests that automatic
function definition reduces the computation effort needed
to solve a problem and produces a simpler and more
understandable overall solution.
#5 Genetics-Based Machine Learning in Rule-Based and
Neural Systems
Professor Robert E. Smith
Department of Engineering Science and Mechanics
The University of Alabama, Tuscaloosa, AL
This tutorial covers the application of genetic algorithms
(GAs) in machine learning. Machine learning is introduced
in the framework of control, with an emphasis on
reinforcement learning, where the system must learn
through a exploration. A brief overview of GAs is also
provided. Given this background, the tutorial discusses
rule-based, neural, and fuzzy techniques that utilize GAs.
A rule-based technique, the learning classifier system
(LCS), is shown to be analogous to a neural network. The
integration of fuzzy logic into the LCS is also discussed.
Research issues related to GA-based learning are
overviewed. The application potential for genetics-based
machine learning is discussed.
#6 Genetic Algorithms:
Theoretical Foundations and Experimental Evaluation
Professor Darrell Whitley
Computer Science Department
Colorado State University, Fort Collins, CO
The principle of hyperplane sampling will be examined, as
well as exact theoretical models of a canonical genetic
algorithm. Other topics include: deception, remapping
hyperspace, stochastic hill climbing versus hyperplane
sampling and the case against gray coding for test
functions. Holland's schema theorem and the K-arm bandit
analogy will be reviewed and critiqued. Alternative forms
of the genetic algorithm such as Genitor, CHC, Evolution
Strategies and parallel genetic algorithms will be
reviewed. The practical implications of the existing
theory will be explored with respect to implementing and
applying genetic algorithms to complex problems. Examples
are given where simple theoretical insights result in
improved search on problems of more than 500 variables.
#7 An Introduction to Fuzzy Logic
Professor James Bezdek
Department of Computer Science
University of West Florida, Pensacola, FL
This tutorial begins by developing the basis for fuzzy
models. The first hour starts with a discussion of
uncertainty in models and its importance for system
design. Membership functions and fuzzy set operations are
defined. We pose and answer some basic questions about
fuzzy models - e.g., where do they come from? how are they
evaluated? how do they compare with probability models?
The second hour presents two applications vignettes. The
first considers stabilization of the simple inverted
pendulum. We compare the classical (linear feedback) and
fuzzy control approaches, and discuss design issues such
as tuning and stability. The second application area is
segmentation of image data. Several approaches based on
fuzzy and neural models are presented and compared.
#8 Fuzzy Sets in Constraint Satisfaction
Dr. Didier Dubois
Institut de Recherche en Informatique de Toulouse
Universite Paul Sabatier, Toulouse Cedex - France
Constraint-directed search is a very general and powerful
methodology for problem solving, which is particularly
adapted to finite domains involving high combinatorial
complexity. The aim of this tutorial is to show that
fuzzy set theory and constraint satisfaction can be easily
and usefully put together. The tutorial will describe the
approach pioneered by Bellman and Zadeh for the modeling
of fuzzy constraints, and point out the difference between
a fuzzy constraint and an objective function, address
how to imbed flexible constraint satisfaction in Zadeh's
calculus of fuzzy relations, whose aim is to propagate
preference in constraint networks, and review in detail
the applications of fuzzy constraint satisfaction in the
field of production research, and especially job-shop
scheduling. The fuzzy methodology will be compared to
knowledge-based job-shop scheduling techniques that come
from Artificial Intelligence.
#9 Fuzzy Logic Applications to Artificial Intelligence
and Intelligent Control Systems
Dr. Enrique H. Ruspini
Artificial Intelligence Center
SRI International, Menlo Park, CA
We present first fuzzy logic as a methodology concerned
with the representation and analysis of vague and
uncertain aspects of reality. Using a unified model of
approximate-reasoning methods, we discuss the nature of
fuzzy-logic methods and compare them with other
uncertainty-modeling techniques such as probabilistic
reasoning.
Using this model, we also show that fuzzy logic is a
sound deductive technique relying on the notions of
utility and preference. Based on such a characterization,
we present an emerging set of procedures for the
development and analysis of fuzzy models. Problems such
as the derivation of possibility distributions, their
interpretation, the representation of vague knowledge, the
integration of multiple conflicting objectives, and the
explanation of planning and control choices are handled in
this framework by means of sound procedures rooted on
logical concepts and principles.
We illustrate the nature of these techniques by means
of examples of their application to the development of
intelligent devices and systems. In particular, we
focus on the architecture and operation of the motion
controller for SRI's Autonomous Mobile Robot, Flakey.
#10 Fuzzy Logic in Computer Vision
Professor James M. Keller
Electrical and Computer Engineering Department
University of Missouri-Columbia, Columbia, MO
Computer vision is the study of theories and algorithms
for automating the process of visual perception. This
involves tasks such as noise removal, smoothing, and
sharpening of contrast; segmentation of images to isolate
objects and regions and description and recognition of the
segmented regions; and finally interpretation of the
scene. The purpose of this tutorial is to give an
overview of the fuzzy set theoretic approach to computer
vision. The applications of fuzzy set theory in computer
vision in the areas of image modeling, preprocessing,
segmentation, boundary detection, object/region
recognition, and reasoning will be discussed.
Techniques presented are demonstrated on real imaging
problems.
#11 Fuzzy Neurocomputations
Professor Witold Pedrycz
Dept. of Electrical and Computer Eng.
University of Manitoba, Winnipeg
Fuzzy neurocomputations as realizing the paradigm of
distributed computations integrate essential learning
capabilities of neural networks with the schemes of
explicit knowledge representation stemming from the
mechanisms of fuzzy sets. This tutorial will address the
issues of constructing, testing, and utilizing fuzzy
neural networks. The cornerstone of fuzzy neural networks
is that their processing elements (neurons) are
constructed with the aid of logical operations available
in the theory of fuzzy sets. Each neuron, as completing
logical operations on the input stimuli, conveys its own
clearly visible semantics. The two classes of neurons will
be studied. The first category of the neurons embraces
aggregation units, while the other one includes
referential operations. The studies of the learning
algorithms applied to the network will include both the
modified gradient-like optimization methods as well as
schemes of genetic optimization. Those latter can be
stratified as they pertain equally well to the structure
of the network, types of the neurons, and the character of
the individual connections. Various applications of the
networks will be also outlined including the utilization
of the networks in designing fuzzy controllers.
#12 Fuzzy Data Analysis
Professor Dr. Dr.h.c.Hans-Jurgen Zimmermann
Professor of Operations Research
RWTH Aachen, Aachen, Germany
This tutorial begins with definitions of basic
terminology. Following this,we discuss methods and
techniques for fuzzy data analysis. Tools discussedwill
include algorithms and software for fuzzy clustering,
decision models that use fuzzy inferencing techniques and
approaches based on combinationsof neural networks and
fuzzy models. The tutorial will illustrate thesetechniques
by discussing applications that include quality control,
imagesegmentation, fault diagnosis and petrochemical
design.
#13 Applications of Neural Networks to Virtual Reality
Professor Thomas P. Caudell
Dept. of Electrical Engineering and Computer Engineering
University of New Mexico, Albuquerque, NM
The objective of this tutorial is to first introduce the
topic of virtual reality and then to show where neural
networks are contributing to this technology. Virtual
Reality (VR) is a form of advanced human-computer
interface technology that embodies a sense of immersion,
interactivity, navigation, and exploration of computer
generated virtual worlds. A relative of VR is Augmented
Reality (AR), where the user remains immersed in the real
world with only small amounts of data being presented. VR
typically involves opaque head-mounted displays that show
only computer generated graphics. AR uses see-through
head-mounted displays that show mostly the real world with
small amounts of computer generated graphics overlaid on
real world objects. There are many technological
challenges left to solve before VR and AR are practical.
Neural networks offer solutions to some of these
challenges, This tutorial will introduce VR and AR
technologies and applications, introduce the classes of
neural networks to be discussed, and illustrate the
application of neural networks to this field with
examples.
#14 Hybrid Systems: Neural, Symbolic, and Fuzzy
Professor Lawrence O. Hall and Professor Abraham Kandel
Computer Science and Engineering Department
University of South Florida, Tampa, FL
Neural networks and expert systems are complementary
approaches to knowledge representation and decision
making. This tutorial concentrates on hybrid systems
which incorporate neural networks to tune expert system
knowledge or will work in concert with an expert system to
solve a problem. The basic concepts underlying hybrid
systems are clearly outlined. The tutorial examines the
question of how to incorporate knowledge into a neural
network and whether symbolic information can be extracted
from a trained neural network. The tutorial examines the
use of fuzzy logic in the neural network expert system
mix. This includes hybrid neuro fuzzy systems. Examples
will be given that show hybrid systems, properly designed,
provide systems more powerful than any of the components
used in a stand-alone fashion.
#15 Basics of Building Market Timing Systems:
Making Money with Neural Networks
Casimir C. Klimasauskas
NeuralWare, Inc.
Pittsburgh, PA
This tutorial will cover the basic principles for building
successful financial market timing systems. Are the
markets predictable? This is the foundation on which this
talk is built. Identifying which markets and when they
are predictable is the first step toward developing a
successful system. In general, the objective of
developing a neural network trading system is to make
money. Building a system which meets the objectives is
the next step and primary focus of this tutorial. The
technological measures on which most neural network
technology is built often fail to maximize system
objectives. Various approaches to addressing these
issues will be discussed. This includes what to predict,
how to modify standard neural paradigms to enhance
ultimate performance, selection of train, test and
verification sets, and data pre-processing. An example of
a system developed on recent data will be used to
illustrate the various issues in the talk.
#16 Practical Applications of Neural Network Theory
Dr. Robert Hecht-Nielsen
HNC, Inc
San Diego, CA
Neural network theory has advanced significantly over the
past five years. In this tutorial, theoretical advances
in the areas of universal approximation, learning and
convergence, curse of dimensionality exorcism, and error
problem-solving will then be described, with an emphasis
on how our practical efforts can be guided by this
theoretical knowledge. Special emphasis will be given to
the topic of which types of problems neural networks are
good at solving and how to select the proper neural
network architecture for a problem. The tutorial is aimed
at those with at least basic familiarity with neural
network architectures and applications. No knowledge of
theory is presumed and no mathematics beyond elementary
calculus andlinear algebra will be used.
#17 Computational Studies of Biological
Neural Networks: Introduction and Applications
to Vision and Sensory-Motor Control
Professor Paolo Gaudiano
Department of Cognitive and Neural Systems
Boston University, Boston, MA
This tutorial introduces an interdisciplinary approach to
the study of computational neural models for uncovering
the functional designs that underlie human and animal
learning and performance. Through a combination of
psychological, physiological, mathematical and
computational notions, the presentation will show how
simple networks of neurons can develop useful functional
properties in response to a rapidly changing and
unpredictable environment. Next, the presentation will
illustrate how these fundamental neural network modules
can be embedded into more elaborate networks that exhibit
complex adaptive behavior,. It will then be shown that
the same fundamental modules serve as building blocks for
other neural network models that can explain biological
function and at the same time provide novel technologies
for practical applications. The presentation will focus
on two examples: one model of low level vision explains
how the vertebrate retina rapidly adjusts its sensitivity
over an enormous range of illumination, a useful property
for artificial vision systems; the other model describes
adaptive sensory-motor control in humans and animals, and
has been applied successfully to visually-guided
navigation of mobile robots.
#18 Learning Algorithms In Neural Networks
Professor Jacek M. Zurada
Computer Science and Engineering
University of Louisville, Louisville, KY
Learning is a fundamental property of networks acquiring
computational intelligence. Learning can be understood as
a change in behavior brought about by experience. In
neural networks learning takes the form of approximation
of relationships from data, or the form of encoding
desired equilibria. This tutorial reviews basic concepts
of supervised and unsupervised learning of most important
neural network architectures. The tutorial stresses the
visualization of learning in both pattern and weight
space. It demonstrates links between various methods of
network adaptation schemes. The material presented is
addressed to persons interested in pursuing independent
research/study/NN modeling who are also seeking
understanding of concepts underlying computational
properties of neural networks.
- ---
IEEE ICNN '94 Program Committee Chairman
Dennis W. Ruck Air Force Institute of Technology
druck@afit.af.mil Wright-Patterson AFB, Ohio
(NeXTmail Welcome)
------------------------------
End of Neuron Digest [Volume 13 Issue 23]
*****************************************
