- Presenter:
- Special guest star: Rick Kazman
- Title: The Architecture Business Cycle and its effects on the WWW
- Subtitle: Mommy, Where do Software Architectures Come From?
- Abstract:
- Abstract: This talk will examine the organizational, commercial, and
professional influences on a software product, and will discuss how
these influences have shaped the software that runs the World-Wide Web.
6. General Discussion Items
- Lab tours for new graduate students.
7. Action List
- University of Toronto visit: November 2
8. Director's Meeting
9. Seminars
The Institute for Computer Research (ICR)
Presents a Seminar on
"Interactive Voice Servers Over the Telecom Network"
by: Professor Gerard Chollet
of: Centre National de la Recherche Scientifique
Ecole Nationale Superieure des Telecommunications-Signal
Paris, France
Date: Wednesday, September 4, 1996
Time: 3:30 p.m.
Place: William G. Davis Computer Research Centre, Room 1304
Abstract:
A lot of 'useful' (and some useless) information is getting more
accessible over the TELECOM network (INTERNET, AudioTex, Banking,
Travel, Tourism ...). In many situations the most convenient
(and sometimes the only) way to access it is Voice I/O.
The use of the telephone on the other hand could be very frus-
trating or distracting when the intended recipient of the call is
absent or busy or does not want to answer the phone . Voice mail
systems are an answer to this problem. They are becoming more
and more user friendly and include such features as setting up
appointments, speaker verification, etc.
This talk will review some to the speech technology necessary to
fully develop this market. Among the issues are: Language-
independent flexible vocabulary approach to speech recognition,
robustness to the environment (noise, channel distortions, etc.),
dialectical variations, speaker verification, adaptation to new
speakers, speech translation, among others. Interactive Voice
Servers developed in France and Switzerland will be described and
evaluated.
Biography:
Gerard Chollet's education was centred on Mathematics (DUES-MP),
Physics (Maotrise), Engineering and Computer Sciences (DEA) until
the doctoral level. He studied Linguistics, Electrical and Com-
puter Science at the University of California, Santa Barbara and
was granted a Ph.D. in Computer Science and Linguistics.
Professor Chollet joined the Centre National de la Recherche
Scientifique in 1978 at the Institut de Phonetique in Aix en Pro-
vence. His main research interests are in phonetics, automatic
speech processing, speech dialogue systems, multimedia, pattern
recognition, digital signal processing, speech pathology and
speech training aids.
Everyone is welcome. Refreshments served.
DEPARTMENT OF COMPUTER SCIENCE
UNIVERSITY OF WATERLOO
SEMINAR ACTIVITIES
MASTERS THESIS PRESENTATION
- Thursday, September 5, 1996
J.Y. Eric Giguere , Graduate Student, Department of
Computer Science, University of Waterloo will speak on
``A Model for Database Support in Rapid Application
Development Environments''.
TIME: 2:30 - 3:30 p.m.
ROOM: Davis Centre Room DC3301
ABSTRACT
This thesis describes a model for database support in
rapid application (RAD) environments. With this model,
users who have very little database expertise can
create interactive database applications requiring a
minimal amount of coding while also minimizing or
eliminating the use of user-modified generated code.
The University of Waterloo
200 University Avenue
Waterloo, Ontario
The Institute for Computer Research (ICR) and the Department of Systems
Design Engineering
Present a Joint Seminar on
"Global Optimization Methods for Neural Network Training"
by: Professor Benjamin W. Wah
of: Department of Electrical & Computer Engineering
University of Illnois at Urbana-Champaign
Urbana, Il
Date: Friday, September 6, 1996
Time: 3:00 p.m.
Place: William G. Davis Computer Research Centre, Room 1302
Abstract:
In this talk, we examine various supervised learning methods for
training feed-forward neural networks. In general, such learning
can be considered as a nonlinear global optimization problem in
which the goal is to minimize the nonlinear error function that
spans the space of the weights. We survey various global optimi-
zation methods suitable for neural-network learning, and propose
the NOVEL method, a Nonlinear global Optimization method Via
External Lead for nonlinear optimization and neural network
learning. By combining global and local searches, we show how
NOVEL can be used to find a good local minimum in the error
space. Our key idea is to use a user-defined trace function that
pulls the search out of a local minimum without having to restart
it from a new starting point. Using five benchmark problems, we
compare NOVEL against some of the best global optimization algo-
rithms and demonstrate its superior improvement in performance.
Biography:
Benjamin W. Wah received his Ph.D. degree in computer science
from the University of California, Berkeley, CA, in 1979. He is
currently a Professor in the Department of Electrical and Comput-
er Engineering, Coordinated Science Laboratory, and Beckman In-
stitute of the University of Illinois at Urbana-Champaign, Urba-
na, IL. Previously, he had served on the faculty of Purdue
University (1979-85), as a Program Director at the National Sci-
ence Foundation (1988-89), as Fujitsu Visiting Chair Professor of
Intelligence Engineering, Tokyo University (1992), and McKay
Visiting Professor of Electrical Engineering and Computer Sci-
ence, University of California, Berkeley (1994). Dr. Wah is
Editor-in-Chief of the IEEE Transactions on Knowledge and Data
Engineering, and serves on the editorial boards of several jour-
nals. His current research interests are in the areas of paral-
lel and distributed processing, computer networks, neural net-
works, evolutionary computation, knowledge engineering, and non-
linear optimization.
Everyone is welcome. Refreshments served.
10. Lab Cleanup (until 12:30 or 5 minutes)
Meeting
Agenda