Meeting
Agenda
May 19, 1999
-
Location:
-
DC1304
-
Time:
-
1:30 p.m. (???)
-
Chair:
-
Liddy Olds
Member
List
1. Adoption of the Agenda - additions or deletions
2. Coffee Hour
-
Coffee hour this week:
-
Josée Lajoie
-
Coffee hour next week:
-
Anyone?
3. Next meeting
THE FOLLOWING IS SUBJECT TO CHANGE!
-
Date:
-
Wednesday, May 26, 1999
-
Location:
-
DC1304
-
Time:
-
1:30 p.m.
-
Chair:
-
Viet-Tam Luu
-
Technical presentation:
-
-
Mike McCool (May 26)
4. Forthcoming
-
Chairs:
- Mark Riddell (June 2)
- Ian Stewart (June 9)
- Daming Yao (June 16)...has a class during lab meeting, if we keep lab meeting time as it has been
-
Tech Presenters:
- Liddy Olds (June 2)
- Mark Riddell (June 9)
- Ian Stewart (June 16)
5. Technical Presentation
- Presenter:
- Marryat Ma
- Title:
- The Direct Manipulation of Pasted Surfaces
- Abstract:
- In this talk, I will discuss a method for the direct manipulation of
tensor product B-spline surfaces. This method is based on Bartels and
Beatty's technique for direct manipulation of spline curves. I will
explain how this method can be adapted to allow the direct
manipulation of a surface in a pasting hierarchy. Finally, I will
give three techniques for directly manipulating composite pasted
surfaces.
6. General Discussion Items
- CGL lab meeting to be rescheduled... Bill has a conflict. Possible
times: 2:30 W (if we can switch with the Shosin group), 12:30 W (but
Bill can't always make it), 1:30 Th. Until further notice, however,
the CGL lab meeting will be held next week at the usual time.
- Blair: Lab duties to be revised. Documentation to be updated; please
check page for your lab duty to make sure it's current.
- New machine in lab, an O2. Name will be monodon, and should be
ready for general use in a week or so.
-
Room change: Due to an upcoming ICR Short Course, we have to move your
following booking for room DC 1304 to room DC 1302 at the same time frame:
Wednesday June 9, 1999 1:30 PM to 2:30 PM
7. Action List
8. Director's Meeting
9. Seminars
** EVENT NOTICE **
Statistics & Actuarial Science
Seminar
Thursday, 20 May 1999 at 3:30PM
Math & Computer, Room 5158
-------
Longitudinal analysis of ordinal responses
Lars Endahl
Visiting Scholar - National Institute of Occupational Health, Denmark
ABSTRACT
In many researchers' opinion self-rated pain is the most natural
measure of musculoskeletal disorders. However, self-rated pain can
only be assessed from the trial subjects' own statements, and there is
at present no gold standard for how to make such measurements. In the
PRIM Study, which is a 3-year follow-up study on work-related
musculoskeletal disorders, self-rated pain is measured through
questionnaires, where the average trouble in the upper limbs during
the last 3 months is scored on a 10-point scale.
Such scale responses are usually analysed by logistic regression after
dichotomizing the scale or by a proportional odds model (or the like)
on the full scale (McCullagh 1980). The drawback to the former
approach is that valuable information about the responses is lost,
whereas the drawback to the latter approach is that the assumptions
implied by the proportional odds model are rarely fulfilled when the
response can take a large number of different values.
I propose a quasi-likelihood approach with robust variance estimation
to analyse self-rated pain scored on a 10-point scale. The mean value
structure is parameterized logit linear; the variance structure
resembles a binomial variance with overdispersion. The
quasi-likelihood method requires only a minimum of assumptions, so the
approach applies in general to ordinal interval-scales. Furthermore,
the method is easily extended to a longitudinal set-up, either by use
of generalized estimating equations or by a generalized linear mixed
model. The proposed method is employed on data from the PRIM study.
The talk will be followed by coffee and cookies in MC 6123. ALL WELCOME!
DEPARTMENT OF COMPUTER SCIENCE
UNIVERSITY OF WATERLOO
SEMINAR ACTIVITIES
MASTER'S THESIS PRESENTATION
- Tuesday, May 25, 1999
Ivan Bowman, graduate student, Dept. of Comp. Sci.,
Univ. of Waterloo, will speak on "Architecture
Recovery for Object-Oriented Systems"
TIME: 1:30-2:30 p.m.
ROOM: DC 1304
ABSTRACT
Recent research has suggested that a software
architecture can be used to record, validate, and
transfer developer understanding of high-level system
abstractions. In order to create architectural
documentation and verify that it is correct, we can use
an architecture recovery process that creates a
software architecture document from a system's
implementation.
In this thesis, we examine how architecture recovery
can be used to create software-architecture documents
for large, object-oriented systems. We have defined a
formal basis for reconstructing architectures based on
Tarski relational algebra, implemented fact extractors
that populate this model based on a system's
implementation, and evaluated our process with case
studies of large object-oriented systems.
We have shown that we can use our system in a semi-
automated process to reconstruct the software
architecture of large object-oriented systems. If
incorporated into a development culture, our
architecture-recovery process would help preserve the
value of software systems.
DEPARTMENT OF COMPUTER SCIENCE
UNIVERSITY OF WATERLOO
SEMINAR ACTIVITIES
AlGORITHMS AND COMPLEXITY SEMINAR
- Friday, May 21, 1999
Igor Benko, graduate student, Dept. Comp. Sci., Univ.
Waterloo, will speak on ``ECF processes''
TIME: 11:00 a.m. -12:00 noon
ROOM: DC 1304
ABSTRACT
ECF processes are a simple formal framework that we use
to model networks of communicating processes. The key
feature of ECF processes is the use of enhanced
characteristic functions (ECFs) in order to attach
labels to sequences of communication actions. We use
these labels to denote various behavioral properties of
a process, such as safety or progress guarantees. The
general ECF model does not rely on a particular set of
labels. Rather, we require that the labels satisfy an
number of simple properties, which we use to define a
number of operations on ECF processes.
In this talk we present the general ECF model and we
illustrate its application to studying safety of
networks of processes.
** EVENT NOTICE **
Combinatorics and Optimization
Seminar
Wednesday, 19 May 1999 at 3:30PM
Math & Computer, Room 5136
-------
Informal Graduate Student Seminar - Symbolic Derivatives of Matrix Functions
Serge Kruk
Dept. of Combinatorics & Optimization, University of Waterloo
ABSTRACT
For a complete abstract, please visit our web-site.
For Additional Information, Contact:
URL: http://www.math.uwaterloo.ca/CandO_Dept/Seminars/upcoming.html
** EVENT NOTICE **
Combinatorics and Optimization
Seminar
Thursday, 20 May 1999 at 10:30AM
Math & Computer, Room 5136
-------
Undergraduate Research Student Seminar - Prof. Godsil's -- Knots and Brackets -
Mr. Eisen's -- Visual Cryptography
Prof. Chris Godsil & Mr. Phil Eisen
Dept. of Combinatorics & Optimization, University of Waterloo
ABSTRACT
For a complete abstract, please visit our web-site.
** EVENT NOTICE **
Combinatorics and Optimization
Colloquium
Friday, 21 May 1999 at 3:30PM
Math & Computer, Room 5158
-------
Tutte Colloquium - Matroid 4-connectivity
Prof. Jim Geelen
Dept. of Combinatorics & Optimization, University of Waterloo
ABSTRACT
For a complete abstract, please visit our web-site.
------------------------------------------------------------------------------
For Additional Information, Contact:
URL: http://www.math.uwaterloo.ca/CandO_Dept/Seminars/upcoming.html
** EVENT NOTICE **
Computer Science
Seminar
Tuesday, 25 May 1999 at 10:30AM
DC 1304
-------
Toward Continuous Program Understanding
Ken Wong
Dept. Computer Science, University of Victoria, British Columbia
ABSTRACT
Software must evolve over time or it becomes useless. Much of
software production today is involved not in creating wholly new
code but in maintaining existing code. Consequently, the
understanding of this code and the communication of that
understanding are important aspects of ensuring long-lived, useful
software.
Unfortunately, these systems may be (or become) poorly documented.
Also, these systems typically become more complex and difficult to
understand over time. This phenomenon is a natural one for any
large software system that undergoes change to meet user needs. Thus,
the continuous understanding of existing code is an important
problem in preserving and extending the meaningful life of a
software system.
To address this problem, one approach is to use reverse engineering
tools to produce system abstractions that describe the high-level
structure of the software. However, many of these tools have met
limited practical success. For example, they have difficulties
adapting to the particular code, the application domain, the
development environment, the change process, or the comprehension
strategies used by the programmer.
In this talk, I will discuss these difficulties and my approach at
addressing them. One goal of this research is to develop effective
tools to allow continuous, incremental reverse engineering for
software evolution. Many of the lessons learned, however, are
applicable to other software engineering tools.
WebNotice for: uw.talks
Events For 15-May-1999 to 21-May-1999
For additional information see:
http://www.math.uwaterloo.ca:80/wreg/notices_main.pl
--------------------------------------------------
Monday, 17 May 1999 at 1:30PM
Institute for Computer Research - ICR Seminars/Lectures: Seminar
Title: "A Survey of Holonic Systems in Intelligent Manufacturing and
Robotics"
Speaker: Dr. William A. Gruver, Department of Engineering Science, Simon
Fraser University
Location: Davis Centre Room 1304
Abstract:
Computer integrated manufacturing with its relatively inflexible
structure of "flexible" manufacturing cells is inadequate for the
majority of industries because of rapidly changing markets, smaller
batch sizes, and increased demands for customization of products.
Future systems must include a variety of integrated subsystems,
including coordination scheduling, resource allocation, information
distribution, schedule optimization, control interfaces, planning
interfaces, and document management. The next generation of
manufacturing systems must involve humans, machines, and software
modules interacting in dynamically-formed virtual clusters to form
reconfigurable, extensible and scaleable systems. This seminar
presents technologies to meet these needs being developed by students
and staff of the Intelligent Robotics and Manufacturing Systems
Laboratory in collaboration with partners of the Holonic Manufacturing
Systems Consortium. Applications to manufacturing scheduling and
execution, automated assembly, and robotic handling will be described.
The research is based on the use of cooperative agents, called holonic
systems, and organized in hardware or software components that each
independently handle a small set of specialized tasks and cooperate to
achieve system-level goals. This partitioning into small components
allows a high degree of flexibility since systems can be easily
reconfigured and extended at run time.
Biography
William A. Gruver is Professor of Engineering Science at Simon Fraser
University in British Columbia, and head of the Intelligent Robotics
and Manufacturing Systems Laboratory. He holds the PhD, MSEE, and BSEE
degrees from the University of Pennsylvania, and the DIC from the
Imperial College of Science and Technology. In 1995 he was elected an
IEEE Fellow for his leadership of major international programs in
robotics and manufacturing systems engineering. Dr. Gruver is the
author of 140 technical articles and 3 books. His current research
interests are intelligent planning and coordination scheduling,
grasping by multifingered hands, and sensor-based robot control.
Currently, he serves as an associate editor of the IEEE Transactions
on Systems, Man and Cybernetics, IEEE Transactions on Control Systems
Technology, and Robotics and Autonomous Systems. He is a Program
Chair of the 1999 IEEE Conference on Systems, Man and Cybernetics to
be held in Tokyo during October 1999 and General Chair of the
NAFIPS/IFSA International Conference on Fuzzy Systems and Soft
Computing to be held in Vancouver during July 2001. He is an active
participant and workpackage leader of the Holonic Manufacturing
Systems Consortium, a major international project with 50 companies
and research laboratories developing cooperative agent technologies
and their applications to intelligent manufacturing.
Invited by Dr. Mohamed Kamel, Department of Systems Design Engineering
Everyone is welcome. Refreshments served.
ICR gratefully acknowledges the support of its Corporate Members for
this Seminar Series: Bell Canada, DALSA Inc. and NCR Canada Limited.
Visit our web site at http://icr.uwaterloo.ca for the latest news on
events sponsored by ICR.
Talk Sponsored by: Pattern Analysis and Machine Intelligence Group
----------------------------------------
Monday, 17 May 1999 at 3:30PM - ** MOVED TO NEW TIME/DATE (see below) **
Combinatorics and Optimization: Seminar
Title: "Cryptography Seminar - The Special Number Field Sieve
Revisited"
Speaker: Prof. Oliver Schirokauer, Department of Mathematics, Oberlin
College
Location: ** MOVED TO ** Friday, 14 May 1999 at 1:30PM
Abstract:
For a complete abstract, please visit our web-site.
For additional Information, Contact:
URL:
http://www.math.uwaterloo.ca/CandO_Dept/Seminars/upcoming.html
----------------------------------------
Monday, 17 May 1999 at 3:30PM
Computer Science - SCG-Symbolic Computation Group: Seminar
Title: "Symbolic-numeric Techniques for Polynomial Systems"
Speaker: Hans J. Stetter, Tech. Univ. Vienna, Austria
Location: DC 1304
Remarks: Co-Sponsored by the Scientific Computation Group
Abstract:
For good reasons, systems of linear equations with numerical
(not symbolic) coefficients are generally solved numerically (i.e.
in floating-point arithmetic), even if the rational data are
considered as exact; if (some) data are of limited accuracy there is
no ``exact solution'' anyway. The same situation prevails in
polynomial algebra (for multivariate systems of polynomial
equations), with an additional incentive: Even with exact
rational coefficients, there is generally no rational solution. On
the other hand, one wants to preserve the algebraic structure of
the problem.
We will consider various fundamental aspects of this situation
from principal and algorithmic points of view:
- a-posteriori appraisal of the quality of approximate
zeros, independently of how they have been obtained;
- representation singularities in polynomial ideals and their
(local) elimination, a prerequisite for numerical methods;
- reduction of a 0-dimensional polynomial system to a matrix
eigenproblem, the key to efficient numerical algorithms for
polynomial equations.
----------------------------------------
Wednesday, 19 May 1999 at 11:30
The infraNET Project: Colloquium
Title: "Canada's Smart Communities"
Speaker: David Johston, President-Designate, University of Waterloo
Location: Federation Hall
Remarks: ** Pre-paid Registration: $25 ** At-the-Door: $40 **
Registration is being handled by Communitech. ** Register via E-mail
chris@communitech.org or by fax 519-888-7007.
Abstract:
Smart Communities will showcase and promote information and
communication technologies so they benefit all Canadians. This
presentation will highlight recommendations to the federal government
of the Blue Ribbon Panel on Smart Communities. The Smart Community
initiative is a promising vision, one of excellence, innovation and
lifelong learning. Smart Communities will ensure that their citizens
are able to participate in the constantly evolving knowledge society.
The changing role of broadband community networks, connectivity, and
services as they relate to Smart Communities will also be discussed.
** Registration 11:30 a.m. - 12:00 p.m. ** Lunch 12:00 p.m. - 12:45
p.m. ** Program 12:45 p.m. - 1:30 p.m.
Talk Sponsored by: University of Waterloo in cooperation with The infraNET
Project and Communitech
For additional Information, Contact:
Name: Shirley Fenton
E-mail: infranet@uwaterloo.ca
Phone: 888-4567 ext. 5611
URL: http://infranet.uwaterloo.ca
----------------------------------------
Wednesday, 19 May 1999 at 3:30PM
Computer Science: Master's Thesis Presentation
Title: "High precision computation via successive refinement"
Speaker: Ka-Wai Chan, Graduate Student, Dept. Computer Science,
University of Waterloo
Location: DC 1331
Abstract:
In this thesis we study the use of successive refinement
techniques to obtain performance improvements for problems in
computer algebra. We make use of the facilities available in Maple to
perform the techniques that are developed.
We illustrate our work by performing many time consuming
problems in computer algebra and compare the direct approach with
that of the iterative one. We show that through the replacement
of costly software floating point operations with fast hardware
floating point operation, many problems can be sped up. We
conclude that successive refinement is an excellent way to improve
the performance of computer algebra problems.
----------------------------------------
Thursday, 20 May 1999 at 10:30AM
Combinatorics and Optimization: Seminar
Title: "Undergraduate Research Student Seminar - Prof. Godsil's --
Knots and Brackets - Mr. Eisen's -- Visual Cryptography"
Speaker: Prof. Chris Godsil & Mr. Phil Eisen, Dept. of Combinatorics &
Optimization, University of Waterloo
Location: Math & Computer, Room 5136
Abstract:
For a complete abstract, please visit our web-site.
For additional Information, Contact:
URL:
http://www.math.uwaterloo.ca/CandO_Dept/Seminars/upcoming.html
----------------------------------------
Friday, 21 May 1999 at 11:00AM
Computer Science - Algorithms and Complexity Group: Seminar
Title: "ECF processes"
Speaker: Igor Benko, graduate student, Dept. Computer Science, University
of Waterloo
Location: DC 1304
Abstract:
ECF processes are a simple formal framework that we use to model
networks of communicating processes. The key feature of ECF
processes is the use of enhanced characteristic functions
(ECFs) in order to attach labels to sequences of communication
actions. We use these labels to denote various behavioral properties
of a process, such as safety or progress guarantees. The general
ECF model does not rely on a particular set of labels. Rather, we
require that the labels satisfy an number of simple properties,
which we use to define a number of operations on ECF processes.
In this talk we present the general ECF model and we illustrate
its application to studying safety of networks of processes.
----------------------------------------
Friday, 21 May 1999 at 3:30PM
Combinatorics and Optimization: Colloquium
Title: "Tutte Colloquium - Matroid 4-connectivity"
Speaker: Prof. Jim Geelen, Dept. of Combinatorics & Optimization,
University of Waterloo
Location: Math & Computer, Room 5158
Abstract:
For a complete abstract, please visit our web-site.
For additional Information, Contact:
URL:
http://www.math.uwaterloo.ca/CandO_Dept/Seminars/upcoming.html
----------------------------------------
DEPARTMENT OF COMPUTER SCIENCE
UNIVERSITY OF WATERLOO
SEMINAR ACTIVITIES
COMPUTER SCIENCE SEMINAR
- Wednesday, May 26, 1999
Rudolf Fleischer, Max-Planck-Institut fur Informatik,
Saarbrucken, will speak on ``Alternative models of
computation''
TIME: 3:30-4:30 p.m.
ROOM: DC 1304
ABSTRACT
Computer technology based on silicon is rapidly
approaching its physical limits. Are optical computers,
DNA-computers or quantum-computers really an
alternative to the traditional digital electronic
computer?
** EVENT NOTICE **
Statistics & Actuarial Science
Seminar
Tuesday, 18 May 1999 at 3:00PM
** MOVED TO ** Friday, 28 May
1999 at 3:30
-------
Generalized Resolution and Minimum Aberration for Plackett-Burman and Other
Nonregular Factorial Designs
Boxin Tang
Department of Mathematics, University of Tennessee, Memphis
ABSTRACT
The resolution has perhaps been the most widely used criterion for
comparing regular factorials since it was introduced in 1961 by Box
and Hunter. In this paper, we examine how a generalized resolution
criterion can be defined and used for assessing nonregular fractional
factorials, notably those Plackett-Burman designs. Our generalization
is aimed to capture the projection properties, complementing that of
Webb (1964) whose concept of resolution concerns the estimability of
lower order effects under the assumption that higher order effects are
negligible. Our generalized resolution provides a fruitful criterion
for ranking different designs while Webb's resolution is mainly
useful as a classification rule. An additional advantage of our
approach is that the idea leads to a ntural generalization of minimum
aberration. Examples are given to illustrate the usefulness of the
new criteria.
The talk will be followed by coffee and cookies in MC 6123. ALL WELCOME!
** EVENT NOTICE **
Statistics & Actuarial Science
Seminar
Friday, 28 May 1999 at 3:00PM
Math & Computer, Room 5158
-------
Estimating Adverse Selection Costs from Genetic Testing for Breast and Ovarian
Cancer: The Case of Life Insurance
Dr. Jean Lemaire
Wharton School of the University of Pennsylvania
ABSTRACT
Genetic testing is a concern for insurers if they are prohibited from
using testing results in underwriting. We evaluate the impact of
adverse selection in an insurance market with genetic testing for
breast and ovarian cancer. We estimate increased forces of mortality
resulting from a family history of cancer, or a positive test for a
BRCA mutation. Using a continuous-time discrete-state Markov model,
we estimate costs of adverse selection, assuming various testing and
insurance purchase behaviours. We conclude that, at current testing
rates, adverse selection should be controllable if companies apply
strict underwriting rules, requesting cancer history and onset age for
all first-degree relatives. If companies do not request precise
family information, or are forbidden by law to obtain it, adverse
selection costs would be so large that major disruptions in the
insurance market would result.
The talk will be followed by coffee and cookies in MC 6123. ALL WELCOME!
------------------------------------------------------------------------------
Talk Sponsored by: Institute of Insurance and Pension Research
10. Lab Cleanup