segre@GVAX.CS.CORNELL.EDU (Alberto M. Segre) (02/05/89)
Call for Papers:
Sixth International Workshop on Machine Learning
Cornell University
Ithaca, New York; U.S.A.
June 29 - July 1, 1989
The Sixth International Workshop on Machine Learning will be
held at Cornell University from June 29 through July 1, 1989.
The workshop will be divided into six parallel sessions, each
focusing on a different theme:
Combining Empirical and Explanation-Based Learning (M. Pazzani,
chair). Both empirical evaluation and theoretical analysis have
been used to identify the strengths and weaknesses of
individual learning methods. Integrated approaches to learning
have the potential of overcoming the limitations of individual
methods. Papers are solicited exploring hybrid techniques
involving, for example, explanation-based learning, case-based
reasoning, constructive induction, or neural networks.
Empirical Learning; Theory and Application (C. Sammut, chair).
This session will be devoted to discussions on inductive (also
called empirical) learning with particular emphasis on results
that can be justified by theory or experimental evaluation.
Papers should characterize methodologies (either formally or
experimentally), their performance and/or problems for which
they are well/ill suited. Comparative studies applying
different methodologies to the same problem are also invited.
Learning Plan Knowledge (S. Chien and G. DeJong, co-chairs).
This session will explore machine learning of plan-related
knowledge; specifically, learning to construct, index, and
recognize plans by using explanation-based, empirical, case-
based, analogical, and connectionist approaches.
Knowledge-Base Refinement and Theory Revision (A. Ginsberg,
chair). Knowledge-base refinement involves the discovery of
plausible refinements to a knowledge base in order to improve
the breadth and accuracy of the associated expert system. More
generally, theory revision is concerned with systems that start
out having some domain theory, but one that is incomplete and
fallible. Two basic problems are how to use an imperfect
theory to guide one in learning more about the domain as more
experience accumulates, and how to use the knowledge so gained
to revise the theory in appropriate ways.
Incremental Learning (D. Fisher, chair, with J. Grefenstette, J.
Schlimmer, R. Sutton, and P. Utgoff). Incremental learning
requires continuous adaptation to the environment subject to
performance constraints of timely response, environmental
assumptions such as noise or concept drift, and knowledge base
limitations. Papers that cross traditionally disparate
paradigms are highly encouraged, notably rule-based,
connectionist, and genetic learning; explanation-based and
inductive learning; procedure and concept learning;
psychological and computational theories of learning; and
belief revision, bounded rationality, and learning.
Representational Issues in Machine Learning (D. Subramanian,
chair). This session will study representational practice in
machine learning in order to understand the relationship
between inference (inductive and deductive) and choice of
representation. Present-day learners depend on careful
vocabulary engineering for their success. What is the nature
of the contribution representation makes to learning, and how
can we make learners design/redesign hypotheses languages
automatically? Papers are solicited in areas including, but not
limited to, bias, representation change and reformulation, and
knowledge-level analysis of learning algorithms.
PARTICIPATION
Each workshop session is limited to between 30 and 50
participants. In order to meet this size constraint, attendance
at the workshop is by invitation only. If you are active in
machine learning and you are interested in receiving an
invitation, we encourage you to submit a short vita (including
relevant publications) and a one-page research summary describing
your recent work.
Researchers interested in presenting their work at one of
the sessions should submit an extended abstract (4 pages maximum)
or a draft paper (12 pages maximum) describing their recent work
in the area. Final papers will be included in the workshop
proceedings, which will be distributed to all participants.
SUBMISSION REQUIREMENTS
Each submission (research summary, extended abstract, or
draft paper) must be clearly marked with the author's name,
affiliation, telephone number and Internet address. In addition,
you should clearly indicate for which workshop session your
submission is intended.
Deadline for submission is March 1, 1989. Submissions should be
mailed directly to:
6th International Workshop on Machine Learning
Alberto Segre, Workshop Chair
Department of Computer Science
Upson Hall
Cornell University
Ithaca, NY 14853-7501
USA
Telephone: (607) 255-9196
Internet: ml89@cs.cornell.edu
While hardcopy submissions are preferred, electronic
submissions will be accepted in TROFF (me or ms macros), LaTeX or
plain TeX. Electronic submissions must consist of a single file.
Be sure to include all necessary macros; it is the responsibility
of the submitter to ensure his/her paper is printable without
special handling. Foreign contributors may make special
arrangements on an individual basis for sending their submissions
via FAX.
Submissions will be reviewed by the individual session
chair(s). Determinations will be made by April 1, 1989.
Attendance at the workshop is by invitation only; you must submit
a paper, abstract or research summary in order to be considered.
While you may make submissions to more than one workshop session,
each participant will be invited to only one session.
IMPORTANT DATES
March 1, 1989
Submission deadline for research summaries, extended
abstracts and draft papers.
April 1, 1989
Invitations issued; presenters notified of acceptance.
April 20, 1989
Final camera-ready copy of accepted papers due for inclusion
in proceedings.