Copy Link
Add to Bookmark
Report
Machine Learning List Vol. 4 No. 16
Machine Learning List: Vol. 4 No. 16
Thursday, August 6, 1992
Contents:
Allen Newell
ML93 call for papers
Human Genome Fellowships
Genetic Algorithms Conference Announcement
Lazy Partial Evaluation Implementation
FOCL
PostDoc at Ottawa
The Machine Learning List is moderated. Contributions should be relevant to
the scientific study of machine learning. Mail contributions to ml@ics.uci.edu.
Mail requests to be added or deleted to ml-request@ics.uci.edu. Back issues
may be FTP'd from ics.uci.edu in pub/ml-list/V<X>/<N> or N.Z where X and N are
the volume and number of the issue; ID: anonymous PASSWORD: <your mail address>
----------------------------------------------------------------------
Subject: Allen Newell
Date: Mon, 27 Jul 92 11:21:49 -0400
From: Frank_Ritter@shamo.soar.cs.cmu.EDU
From: The CMU Soar Research Group
Allen Newell, one of the founders of the fields of artificial
intelligence and cognitive science, died July 19 in Pittsburgh. He
was 65.
Newell earned an international reputation for his pioneering work in
artificial intelligence, the theory of human cognition and the
development of computer software and hardware systems for complex
information processing.
Last month he was awarded the National Medal of Science by President
Bush.
Newell's career spanned the entire computer era, which began in the
early 1950s. In computer science, he worked on areas as diverse as
list processing, computer description languages, hypertext systems and
psychologically based models of human-computer interaction.
The fields of artificial intelligence and cognitive science grew in
part from his idea that computers could process symbols as well as
numbers, and if programmed properly would be capable of solving
problems in the same way humans do. In the 1960's (in particular)
Allen and Herb Simon created computer models of human problem solving.
This work was one of the major forces behind the "cognitive
revolution" in psychology.
Throughout his research career his work touched on architectures to
support intelligent action in humans and machines. Since the early
1980s, his research interests were centered on the development of
Soar, a cognitive architecture realized as a software system capable
of solving problems and learning in ways similar to human beings. As
a proposed unified theory of cognition, the goal of Soar is to provide
an underlying structure that would enable a computer system to perform
the complete range of mental tasks. Soar has been in use for the past
five years as a framework for intelligent system design at research
institutions around the world.
A native of San Francisco, Newell received a bachelor's degree in
physics from Stanford University in 1949. He spent a year at Princeton
University doing graduate work in mathematics, and worked for the Rand
Corporation as a research scientist from 1950-61. While at Rand, he met
Nobel Laureate Herbert A. Simon, then a professor of industrial
administration at Carnegie Institute of Technology (CIT), now Carnegie
Mellon University. Their discussions on how human thinking could be
modeled led Newell to come to Pittsburgh so the two could collaborate.
Newell earned a doctor's degree in industrial administration from CIT's
business school in 1957.
Newell joined the CIT faculty as a professor in 1961. He played a
pivotal role in creating Carnegie Mellon's School of Computer Science
and elevating the school to world-class status.
Newell, a professor of psychology and the U.A. and Helen Whitaker
professor of computer science at the time of his death, wrote and
co-authored more than 250 publications, including 10 books. He
co-authored "Human Problem Solving" with Simon in 1972, and
co-authored "The Psychology of Human-Computer Interaction" with two
colleagues in 1983. His most recent book, "Unified Theories of
Cognition," published by Harvard University Press in 1990, is based on
the thesis that tools are at hand that will allow psychologists to
start to develop a unified theory describing many different types of
behavior, instead of building separate theories to cover isolated
aspects, as has long been the practice.
Newell's awards and honors include the Harry Goode Award of the
American Federation of Information Processing Societies (1971); the
A.M. Turing Award of the Association for Computing Machinery, jointly
with Simon (1975); the Alexander C. Williams Jr. Award of the Human
Factors Society (1979); the Distinguished Research Contribution Award
of the American Psychological Association (1985); the Research
Excellence Award of the International Joint Conference on Artificial
Intelligence (1989); the Emanuel R. Piore Award of the Institute for
Electrical and Electronic Engineers (1990); and the Franklin
Institute's Louis E. Levy Medal (1992). He was awarded honorary
doctor degrees by the University of Pennsylvania and the University of
Groeningen in the Netherlands.
Newell was a member of the National Academy of Sciences, the National
Academy of Engineering and the American Academy of Arts and Sciences.
He was the first president of the American Association for Artificial
Intelligence and president of the Cognitive Science Society. In 1987
he delivered the William James Lectures to the Department of
Psychology at Harvard. Those lectures formed the basis for his book,
"Unified Theories of Cognition."
Newell is survived by his wife, Noel and a son, Paul, who lives in
California.
For more information contact:
Dr. Jill Fain Lehman
School of Computer Science
Carnegie Mellon Univesity
5000 Forbes Ave.
Pittsburgh, PA 15213
(412) 268-6246
electronic mail: Jill.Lehman@cs.cmu.edu
[25-Jul-92]
------------------------------
Date: Thu, 30 Jul 92 20:48:20 EDT
From: utgoff%zinc@cs.umass.EDU
Subject: ml93 call for papers
Tenth International Conference on Machine Learning
Amherst, Massachusetts: June 27-29, 1993
Call for Papers
The Tenth International Conference on Machine Learning will be held at the
Amherst campus of the University of Massachusetts during June 27-29, 1993,
with informal workshops on June 30th. Paper submissions are welcome from
researchers in computer science, psychology, statistics, neuroscience, and
anyone else who can offer insight on computational learning mechanisms. The
conference will include presentations of refereed papers and invited talks.
Review Criteria
Each submitted paper will be reviewed by at least two members of the program
committee, and be judged on clarity, significance, and originality. Each
submission should contain new results that have not been published
previously. Authors are encouraged to state explicitly the advance in our
understanding of learning phenomena that their work offers. Each accepted
conference paper will be allotted eight (8) proceedings pages and a thirty
(30) minute oral presentation. Outstanding papers will be selected for
presentation in plenary sessions.
Paper Format for Review
Each submission must be clearly legible, with good quality print. Papers
are limited to a total of sixteen (16) double-spaced pages, formatted with a
twelve (12) point font and one (1) inch margins. Each paper must include the
postal and e-mail addresses of all authors, an abstract, and at least three
keywords indicating the problem area, general approach, and evaluation
methodology.
Requirements for Submission
Authors must submit four copies of their papers to the address below by
January 29, 1993. Double-sided copy is acceptable, but electronic or FAX
submission is not. Notification of acceptance or rejection will be mailed to
the first (or designated) author by March 3, 1993. Camera-ready copy of
accepted papers will be due April 7, 1993. Send conference paper submissions
to:
Prof. Paul E. Utgoff (ML93)
Department of Computer Science
Lederle Graduate Research Center
University of Massachusetts
Amherst, MA 01003
Program Committee
Peter Cheeseman NASA Ames
William Cohen AT&T Bell Labs
Oren Etzioni U. Washington
Usama Fayyad JPL
Douglas Fisher Vanderbilt U.
John Grefenstette NRL
Kristian Hammond U. Chicago
Haym Hirsh Rutgers U.
Robert Holte U. Ottawa
Dennis Kibler U. C. Irvine
Pat Langley NASA Ames
Sridhar Mahadevan IBM Watson
Ryszard Michalski George Mason U.
Tom Mitchell Carnegie-Mellon U.
Ray Mooney U. T. Austin
Katharina Morik U. Dortmund
Stephen Muggleton Turing Institute
Leonard Pitt U. Illinois
Ross Quinlan U. Sydney
Larry Rendell U. Illinois
Paul Rosenbloom ISI / U. S. C.
Stuart Russell U. C. Berkeley
Lorenza Saitta U. Torino
Jude Shavlik U. Wisconsin
Richard Sutton GTE
Devika Subramanian Cornell U.
Kurt Vanlehn U. Pittsburgh
Steven Whitehead U. Rochester
Jan Zytkow Wichita State U.
Informal Workshops
Proposals are invited for informal workshops in areas of interest related to
machine learning. Proposals will be reviewed by members of the organizing
committee in order to provide some overall coordination. However, the
detailed arrangements for the program of each workshop will be the
responsibility of the workshop organizers. Only help with local arrangements
will be provided. Send a two-page proposal to the address above by November
13, 1992, indicating the organizer(s), nature and objective of the proposed
workshop, and the likely number of attendees.
General Conference Information
Send inquiries to ml93@cs.umass.edu.
------------------------------
Date: Mon, 3 Aug 92 19:07:39 -0400
From: Larry Hunter <hunter@nlm.nih.GOV>
Subject: Human Genome Fellowships
These fellowships may be of interest to ML researchers who want to get into
molecular biology domains.
***** begin forwarded message *****
The National Center for Human Genome Research, NIH, would like
to remind interested scientists that fellowships are available
for those interested in pursuing a career in genomic research.
Fellowships are also available to scholars in the humanities
who wish to pursue a program of study under the Ethical, Legal
and Social Implications Program. Fellowships are available
for junior scientists (F32) and senior scientists (F33). Only
U.S. citizens or permanent residents are eligible for these
fellowships.
The next receipt date for applications is September 10. For
those interested in genomic research, please contact Dr. Bettie
Graham at (301) 496*7531. For those interested in ethical,
legal and social implications of genomic research, please contacat
Dr. Eric Juengst at (301) 402*0911.
Bettie J. Graham, Ph.D.
Research Grants Branch
National Center for Human Genome Research
National Institutes of Health
***** end forwarded message *****
------------------------------
Subject: Genetic Algorithms Conference Announcement
Date: Wed, 05 Aug 92 09:11:45 -0600
From: Robert Elliott Smith <rob@comec4.mh.ua.EDU>
Fifth International Conference
on Genetic Algorithms
ICGA93
17-22 July, 1993
University of Illinois at
Urbana-Champaign
PRELIMINARY ANNOUNCEMENT
The Fifth International Conference on Genetic Algorithms (ICGA-93),
will be held July 17-22, 1993 at the University of Illinois at
Urbana-Champaign. This meeting brings together an international
community of researchers and practitioners from academia and
industry interested in algorithms suggested by the processes of
natural evolution. Topics of interest will include the design, analysis,
and application of genetic algorithms in optimization and machine learning.
Machine learning architectures of interest include classifier systems and
connectionist schemes that use genetic algorithms.
Papers discussing how genetic algorithms are related to evolving system
modeling (e.g., modeling of nervous system evolution, computational ethology,
artificial life, economics, etc.) are also encouraged.
A formal call for papers for ICGA-93 will be released in the coming weeks.
In the meanwhile, for further information contact one of the conference
co-chairs,
Dave Schaffer ((914) 945-6168, ds1@philabs.philips.com) or
Dave Goldberg ((217) 333-0897, GOLDBERG@vmd.cso.uiuc.edu) or
the publicity chair,
Rob Smith ((205) 348-1618, rob@comec4.mh.ua.edu)
------------------------------
Date: Thu, 6 Aug 92 12:11:22 -0400
From: Peter Clark <pclark@csi.uottawa.ca>
Subject: Lazy Partial Evaluation Implementation
LAZY PARTIAL EVALUATION: Implementation now available!
======================================================
As presented at the 1992 International Machine Learning Conference (pp82-91),
Lazy Partial Evaluation (LPE) is a speedup learning technique integrating
explanation-based generalisation (EBG) and partial evaluation (PE). From
the EBG viewpoint, LPE is similar to EBG except it also stores (rather than
discards) work exploring proofs which ultimately failed. From the PE viewpoint,
LPE only partially evaluates the domain theory just as much is needed to
solve the current problem. LPE can be significantly faster than PE and EBG
on certain types of problems.
A Quintus Prolog implementation of LPE is now available via anonymous ftp from
ftp.csi.uottawa.ca, with the following highlights:
* only need to know two predicates to use LPE!
lpe_setup(Functor/Arity) % set up for LPE for Goal with F/A
lpe_call(+Goal) % evaluate Goal, doing LPE at same time
lpe_call/1 behaves similarly to Prolog's call/1, except that it also has
the side-effect of lazily partially evaluating the domain theory just as
much as is needed to prove Goal.
* user-switchable to work in EBG or PE modes also
* implementation illustrates the similarities between LPE, PE and EBG
* documented
* it works! (well it did last time, anyway...)
* simple walk-through demo comparing LPE, EBG and PE on a toy problem
* more sophisticated examples of LPE in use
* tested under Quintus Prolog on Unix workstations
* copy of the ML92 paper (postscript) also available from this ftp source
* it's free!
Peter Clark Ottawa Machine Learning Group
email: pclark@csi.uottawa.ca Department of Computer Science
phone: (613) 564-5427 University of Ottawa
fax: (613) 564-9486 Ottawa, Ontario K1N 6N5, Canada
To download:
unix> ftp ftp.csi.uottawa.ca
Name: anonymous
Password: (your email address)
ftp> cd /pub/lpe
ftp> get lpecode.tar.Z.uue (or lpepaper.ps.Z.uue for ML92 paper)
ftp> quit
unix> uudecode lpecode.tar.Z.uue
unix> uncompress lpecode.tar.Z
unix> tar xvf lpecode.tar
------------------------------
To: ml@pan.ICS.UCI.EDU
Subject: FOCL
From: Michael Pazzani <pazzani@pan.ICS.UCI.EDU>
FOCL is a machine learning system that extends Quinlan's FOIL program
by containing a compatible explanation-based learning component. FOCL
learns Horn Clause programs from examples and (optionally) background
knowledge.
FOCL is now available by anonymous ftp from ics.uci.edu
For details on FOCL, see:
Pazzani, M. & Kibler, D. (1992). The role of prior knowledge in
inductive learning. Machine Learning, 9, 54-97.
It is available in one of two forms:
1. A (binhexed, Compacted) Macintosh application. This is stored in
pub/SaranWrap/KR-FOCL-ES.cpt.hqx
In addition to the machine learning program, this contains a graphical
interface written by Cliff Brunk that displays the search space
explored by FOCL, so it is a useful pedagogical tool.
This application also contains a graphical interface for building rule
bases, so you can ignore the machine learning aspects, and use it as
an expert system shell with the following capabilities:
* A backward-chaining rule interpreter.
* A graphical rule and fact editor.
* Graphical display of the rule base.
* (Simple) Natural language explanation of inferences
* Menu-based facilities for editing rules and adding natural language
translations to rules.
* Optional typing of variables and checking the rule base for type conflicts
* Tracing of rules
The expert system has been used successfully in an undergraduate
laboratory course. A 75 page manual (that should print on any
Macintosh printer) and sample rule bases are included. A minimum of
4MB of memory is recommended for the application.
If you don't have access to ftp, and you want a copy of the Macintosh
application, send 2 800K disks and a stamped self-addressed envelope
to:
Mike Pazzani
ICS Dept
UC Irvine,
Irvine, CA 92717
USA
2. Common lisp source code. See the file pub/SaranWrap/README for
details on ftping the source code. This is portable source code for
the machine learning program only, since the interface depends on
the MAC.
If you use a copy of FOCL, please send mail to pazzani@ics.uci.edu so
we can inform you of upgrades
------------------------------
From: Stan Matwin <stan@csi.uottawa.ca>
Date: Tue, 4 Aug 92 12:37:21 EDT
Subject: PostDoc at Ottawa
The Department of Computer Science, University of Ottawa, invites
applications for a postdoctoral position on the project "Machine
Learning and Text Analysis for Semi-automatic Knowledge Acquisition".
Candidates must have a completed Ph.D., a strong background in Machine
Learning and Natural Language Processing, and a record of research in
at least one of these areas. The position is available for a year,
with a possibility of renewal for another year. The salary is CDN $
27,500/year, plus an excellent benefit package.
The postdoctoral fellow will be an essential member of the project's team,
now consisting of two faculty, a postdoc and a few graduate students, and
will play an active role in all aspects of the project, including the major
research decisions. Duties include participation in the design and
development of a prototype system, in supervision of graduate students who
will work on the project, and in the publications that will arise from the
project.
Inquiries and requests for more details:
Dr. Stan Matwin (stan@csi.uottawa.ca, tel. 613-564-5069)
Dr. Stan Szpakowicz (szpak@csi.uottawa.ca, tel. 613-564-2450).
fax: 613-564-9486
The mailing address:
Department of Computer Science
University of Ottawa
150 Louis Pasteur
Ottawa, Ontario K1N 6N5
Canada
------------------------------
End of ML-LIST 4.16 (Digest format)
****************************************