Machine Learning List: Vol. 15, No. 18
Saturday, October 18, 2003
Contents
Meeting Announcements
CFP: Workshop on Behavior-based User Interface Customization
CFP: ADM03 - The 2nd Australasian Data Mining Workshop
Workshop on Genomic Signal Processing and Statistics
Seventh Workshop on Mining Scientific and Engineering Datasets
ADM03: Paper submission deadline extension
Miscellaneous Announcements
Rob Schapire and Yoav Freund receive Goedel Prize for AdaBoost
CALL FOR NOMINATION: 2004 SIGART Autonomous Agents Research Award
Call for Site Proposals for ICML-2005
The Machine Learning List is moderated. Contributions should be
relevant to the scientific study of machine learning. Please send
submissions for distribution to: [EMAIL PROTECTED] For requests to be
added, removed, or to change your email address, send email to:
[EMAIL PROTECTED]
To keep mailings to a manageable size, please keep submissions brief.
For meeting announcements, do highlight the meeting Web site and the
goals of the event but omit information such as the program committee
and talk schedules. Also, only first calls for papers and change of
deadline announcements will be included. The ML List moderator
reserves the right to omit/edit submissions to meet these criteria.
----------------------------------------------------------------------
From: Tessa Lau <[EMAIL PROTECTED]>
Subject: CFP: Workshop on Behavior-based User Interface Customization
Date: Sat, 4 Oct 2003 10:38:15 -0400
CALL FOR PAPERS
Workshop on Behavior-based User Interface Customization
Jan 13, 2004 at IUI/CADUI 2004, Maderia, Portugal
http://www.research.ibm.com/iui-workshop/
BACKGROUND AND MOTIVATION
End users of today's applications have limited ability to customize
their user interface to fit specific user preferences and
tasks. Although users come in all shapes and sizes, they must make do
with "one size fits all" applications. Furthermore, particular tasks
may require that the user hop between several applications, each with
its own interface characteristics, often not designed to work well
together. User interface customization allows interfaces to be
adapted to particular user preferences, and specialized to the
specific tasks that users need to perform. It may take a variety of
forms, including: changing the layout to hide irrelevant menus and
buttons and highlight frequently-used options; changing font and icon
sizes; designing new views not originally supported by the application
designer; and providing macro operators that automate common
procedures within and across applications.
An emerging area of research focuses on automatically customizing
applications based on observations of user behavior, employing
techniques from machine learning, plan recognition, and the broader AI
community. For example, systems that observe a user's actions may be
able to infer users' tasks and, through this understanding, be able to
adapt user interfaces that will facilitate performance of tedious
tasks or provide assistance with complex procedures. Further examples
of automatic customization include:
* adaptive web sites based on web usage logs
* inferring a user's skill level and adapting the interface
appropriately
* intelligent macros constructed through programming by demonstration
* dynamic interface layout based on common usage patterns
The purpose of this workshop is to bring together researchers with
experience in the field of behavior-based customization, in order to
consolidate our knowledge of current techniques and identify key
research challenges for the future. The workshop will be organized
around a small number of focus questions. We expect the outcome of
the workshop to:
* Identify the required characteristics for a common platform (e.g.,
at the operating system level) for research into behavior-based
customization,
* Define the limits of customization, both practical and theoretical,
* Determine the "sweet spots", where behavior-based UI customization
is mostly likely to be successful, and where existing AI techniques
best apply, and
* Define the open research questions and agenda that need to be
addressed to move the area of behavior-based UI customization
forward.
IMPORTANT DATES
Submission of workshop papers: November 3, 2003
Notification of acceptance: November 24, 2003
Workshop at IUI/CADUI conference: January 13, 2003
For questions and submission details, please contact
[EMAIL PROTECTED] or visit the workshop web site:
http://www.research.ibm.com/iui-workshop/
WORKSHOP ORGANIZERS
* Lawrence Bergman, IBM TJ Watson Research
* Tessa Lau, IBM TJ Watson Research
------------------------------
From: "Simeon J. Simoff" <[EMAIL PROTECTED]>
Subject: CFP: ADM03 - The 2nd Australasian Data Mining Workshop
Date: Sun, 5 Oct 2003 19:41:12 +1000
CALL FOR PAPERS
ADM03 - The 2nd Australasian Data Mining Workshop
8 December 2003, Lakeside Hotel, Canberra
In conjunction with
the 2003 Congress on Evolutionary Computation (CEC 2003)
(http://www.cs.adfa.edu.au/cec_2003/)
Paper submission due 24 October 2003.
Workshop Website: http://datamining.csiro.au/adm03/
The Australasian Data Mining Workshop is devoted to the art and
science of data mining: the analysis of (usually large) data sets to
discover relationships and present the data in novel ways that are
compact, comprehendible and useful for researchers and
practitioners. We are paritcularly seeking from the Data Mining
community updates of research and progress in the local context, new
breakthroughs in data mining applications and algorithms, evolutionary
approaches to Data Mining, and Bioinformatics and Data Mining.
Data mining projects involve both the utilisation of established
algorithms from machine learning, statistics, and database systems,
and the development of new methods and algorithms, targeted at large
data mining problems. Nowadays data mining efforts have gone beyond
crunching databases of credit card usage or stored transaction
records. They have been focusing on data collected in the health care
system, art, design, medicine and biology and other areas of human
endeavour.
Similar to ADM02, this year edition of the workshop aims to bring
together researchers and industry practitioners from different data
mining groups in Australia and the region and to provide a forum for
presenting and discussing their latest research and development in the
area. The workshop will facilitate the cross-disciplinary exchange of
ideas, experience and potential research directions.
Publication
As with the previous workshop (ADM02) all papers will be reviewed by a
panel of international experts and will be published in a proceedings
volume for the conference. Revised papers will be published by
Springer-Verlag as a collected volume in the Lecture Notes in Computer
Science series.
Topics of interest
The major topics of this year workshop include but are not limited to
* Data mining methods and algorithms
* Data cleaning and data linking
* High performance computing and data mining
* Infrastructure for data mining
* Health Data Mining
* Financial data mining
* Spatial and temporal data mining
* Multimedia and web data mining
* Data mining in design
* Data mining in e-Business environments and virtual communities
* Applications - lessons and experiences
* Data mining in Bioinformatics
We encourage submissions of `greenhouse' work, which present early
stages of a cutting-edge research and development. Software
demonstrations are also welcome. The format of the workshop will
accommodate full paper presentations and short presentations about a
work in progress, overview of a data mining group or software
demonstration.
Submission
The length of the submissions is not restricted. We encourage
submissions of 10-15 pages. The first page of your submission should
include the paper title; author name(s) and affiliation, address,
email; keywords; and abstract. Electronic submissions in either PDF,
PS, RTF or Microsoft Word Document format are preferable. Please,
e-mail electronic submissions to [EMAIL PROTECTED] with subject
"ADM03 Submission".
Important Dates
Submission deadline: 24 October 2003
Notification: 3 November 2003
Camera ready copy: 10 November 2002
Workshop day: 8 December 2003
Organisers
Simeon J Simoff, University of Technology, Sydney
email: [EMAIL PROTECTED]
Graham J Williams, CSIRO Canberra
email: [EMAIL PROTECTED]
Markus Hegland, Australian National University
email: [EMAIL PROTECTED]
REGISTRATION
Please, refer to the conference web page for the registration details.
------------------------------
From: Giovanni Parmigiani <[EMAIL PROTECTED]>
Subject: Workshop on Genomic Signal Processing and Statistics
Date: Wed, 08 Oct 2003 15:03:05 -0400
GENSIPS 2004 -- CALL FOR PAPERS
The Workshop on Genomic Signal Processing and Statistics (GENSIPS
2004) is a cooperating workshop of the IEEE Signal Processing Society,
and will take place on the Homewood campus of The Johns Hopkins
University, May 26-28, 2004.
The aim of this three-day workshop is to provide a forum for
presenting new results on genomic signal processing and statistics for
functional genomics and systems biology and identify potential areas
of research and collaboration between the biological, statistical, and
signal processing communities. One of the main objectives is to
identify new avenues of research, which address modern challenges in
functional genomics, by exploiting potential synergies between signal
processing, statistics and biology and by building on their respective
strengths. Such problem areas might include: signal processing and
extraction of information from microarray images; statistical analysis
of microarray data (classification, gene selection, regulatory network
inference, and clustering); information theoretic approaches to
modeling and analysis of genomic regulatory networks and systems;
signal processing and statistical techniques for the analysis of
protein data and inference of protein networks; and novel
high-throughput hardware/software approaches to genome-scale network
modeling and analysis. This workshop will consist of both invited
sessions and contributed sessions. Invited speakers will give tutorial
talks on the general area of computational functional genomics and
proteomics.
The workshop will be held on the main campus of The Johns Hopkins
University, which is located near downtown Baltimore. It is
financially supported by DARPA, NSF and The Whitaker Foundation, and
is sponsored by The Whitaker Biomedical Engineering Institute and the
Department of Electrical and Computer Engineering of The Johns Hopkins
University.
This call for papers is to solicit contributed papers for the poster
sessions which are expected to be highly interactive. Those interested
should submit a four-page summary describing original work. Final
version of accepted papers will be published in electronic proceedings
which will be distributed by the web and by CD-ROM at the workshop.
Acceptance will be based on quality, relevance and originality.
Participation at the workshop will be limited to 130 attendees.
Registration fee will be kept very low and travel grants will be
offered to selected student participants.
AREAS OF INTEREST:
Digital signal processing and statistical approaches for functional
genomics problems
Digital communications approaches for reverse engineering biological
networks
Data mining and pattern recognition methods for functional genomics
Control theory and systems theory techniques for systems biology
Models for cellular metabolism and inter-cellular signaling
Computational methods for modeling and simulation of biological
regulatory networks
Novel architecture and implementation methods for large-scale
functional genomics
DEADLINES:
January 9, 2004: Four page summaries due
February 9, 2004: Accept/reject notifications sent out
April 26, 2004: Final four-page camera-ready papers due
For detailed submission instructions, please visit the workshop
web page at
http://www.cis.jhu.edu/gensips2004
------------------------------
From: Michael Burl <[EMAIL PROTECTED]>
Subject: Seventh Workshop on Mining Scientific and Engineering Datasets
Date: Fri, 10 Oct 2003 10:32:14 -0600 (MDT)
Seventh Workshop on Mining Scientific and Engineering Datasets
http://www.cs.colorado.edu/~mburl/MSD04/
April 24, 2004
To be held at the Hilton in the Walt Disney World Resort, Lake Buena
Vista, FL in conjunction with the Fourth SIAM International Conference
on Data Mining (SDM 2004)
DEADLINES
October 1, 2003: Call for papers
January 21, 2004: Paper Submission Deadline
February 20, 2004: Acceptance Notification to Authors
March 15, 2004: Camera-ready paper for workshop proceedings
April 24, 2004: Workshop
Although commercial applications of data mining in areas such as
e-commerce, market-basket analysis, text-mining, and web-mining have
taken on a central focus in the KDD community, there is a significant
amount of innovative data mining work taking place in the context of
scientific and engineering applications that is not well-represented
in the mainstream KDD/Data Mining conferences. For example, scientific
data mining techniques are being developed and applied to diverse
fields such as remote sensing, physics, chemistry, biology, astronomy,
etc. In these areas, data mining frequently complements and enhances
existing analysis methods based on statistics and exploratory data
analysis.
On the surface, it may appear that data from one scientific field, say
genomics, is very different from another field, such as
physics. Despite this diversity, there is much that is common across
the mining of scientific data. For example, techniques used to
identify objects in images are very similar, regardless of whether the
images came from a remote sensing application, a physics experiment,
an astronomy observation, or a medical study. Further, with data
mining being applied to new types of data, such as mesh data from
scientific simulations, there is the opportunity to apply and extend
data mining to new scientific domains.
This is the seventh workshop in the series; for information on earlier
workshops, see
http://www.cs.colorado.edu/~mburl/MSD/
The goal of this workshop is to bring together data miners and
scientists who are analyzing science and engineering data from diverse
fields to:
+ share their experiences with other researchers working with similar
data
+ learn how techniques developed in other areas can be applied to their data
+ better understand some of the new techniques being developed in the
scientific data mining community
Topics of interest include, but are not limited to:
+ feature extraction from image and mesh data
+ data fusion techniques for multi-spectral, multi-resolution data
+ data mining system architecture issues
+ application of data mining techniques
+ mining noisy data
+ techniques and applications of computational scientific discovery
+ practical experiences, success stories, and lessons learned
+ dimensionality reduction techniques
+ application of statistical techniques
All papers should be in the context of data from science and
engineering applications. Papers should not exceed 10 pages (including
figures, references, and tables). They should be single spaced, two
column, 10 point, with 1 inch margin all around. An electronic version
of your submission (.pdf or .ps format) should be sent to
[EMAIL PROTECTED] by the deadline given above.
------------------------------
From: "IIS:IIPWM'04 Conference" <[EMAIL PROTECTED]>
Subject: Paper submission deadline extension
Date: Fri, 17 Oct 2003 12:08:31 +0200
INTELLIGENT INFORMATION SYSTEMS 2004 - IIS'04
New Trends in Intelligent Information Processing and Web Mining (IIPWM)
Zakopane, Poland, May 17-20, 2004
Conference Web page: http://iipwm.ipipan.waw.pl
Upon many requests we have received, the deadline for papers'
submission is extended to *** NOVEMBER 3, 2003 ***.
Important deadlines:
http://iipwm.ipipan.waw.pl/2004/dates.html
Scope:
http://iipwm.ipipan.waw.pl/2004/scope.html
Special tracks:
http://iipwm.ipipan.waw.pl/2004/tracks.html
Publication:
http://iipwm.ipipan.waw.pl/2004/publication.html
------------------------------
From: Manfred Warmuth <[EMAIL PROTECTED]>
Subject: Rob Schapire and Yoav Freund receive Goedel Prize for AdaBoost
Date: Tue, 7 Oct 2003 14:16:51 -0700 (PDT)
The Goedel Prize is one of most prestigious prizes in Theoretical
Computer Science (jointly sponsored by EATCS and SIGACT). This is the
first time that a paper in Machine Learning received this award.
See http://sigact.acm.org/prizes/godel for some background information
on the Goedel Prize and a list of past recipients.
Rob Schapire and Yoav Freund received the 2003 prize for their famed
AdaBoost paper. For an announcement see:
http://sigact.acm.org/prizes/godel/2003.html
Background:
Michael Kearns and Les Valiant first defined weak and strong learning
and posed the open problem whether weak learning and strong learning
are the same. In short weak learners must have accuracy only slightly
better than 50% and strong learners must be able to achieve high
accuracy.
In his 1991 Ph.D. thesis from MIT Rob gave the first recursive
construction for combining many weak learners to form a strong
learner. This was followed by Yoav Freund's Ph.D. thesis in 1993,
where he gave a simple flat scheme of combining weak learner by a
majority vote.
After graduating from Santa Cruz, Yoav accepted a job at AT&T Bell
labs in what was one of the strongest machine learning research groups
in the country. Rob was part of that group as well. They combined
ideas and came up with an ``adaptive'' Boosting algorithm (called
AdaBoost - just 10 lines of code) which received a lot of attention in
the Machine Learning and Statistics communities.
Prize winning paper that introduced AdaBoost: "A Decision Theoretic
Generalization of On-Line Learning and an Application to Boosting,"
Journal of Computer and System Sciences 55 (1997), pp. 119-139.
------------------------------
From: Mueller Joerg <[EMAIL PROTECTED]>
Subject: CALL FOR NOMINATION: 2004 SIGART Autonomous Agents Research Award
Date: Wed, 8 Oct 2003 10:02:48 +0200
CALL FOR NOMINATIONS
2004 SIGART Autonomous Agents Research Award
ACM SIGART, in collaboration with the International Conference on
Autonomous Agents, has instituted an annual award for excellence in
research in the area of autonomous agents. Award winners will receive
an honorarium and will be invited to give a presentation at the annual
Autonomous Agents and Multi-Agent Systems (AAMAS) Conference.
This award is specifically intended to recognize researchers whose
current research is influencing the field. Award candidates will
therefore be evaluated based on the quality and signficance of their
research contributions over the last five years. It is expected that
at least some of these contributions should have been reported at one
or more Autonomous Agents or AAMAS conferences. Previous winners of
the SIGART Autonomous Research Award were Nick Jennings (2003), Tuomas
Sandholm (2002), and Katia Sycara (2001).
The award committee is now seeking nominations for next year's award.
If you wish to place a nomination, complete the form below and send it
to Joerg Mueller ([EMAIL PROTECTED]). Nominations are
requested no later than December 31, 2003.
Name of person being nominated: ______________________
Name of person making the nomination: __________________
Please summarize below the reasons why the nominee should be considered
for the award.
------------------------------
From: "Thomas G. Dietterich" <[EMAIL PROTECTED]>
Subject: Call for Site Proposals for ICML-2005
Date: Thu, 9 Oct 2003 12:58:10 -0700
INTERNATIONAL CONFERENCE ON MACHINE LEARNING 2005
CALL FOR SITE PROPOSALS
In the summer of 2005, the International Conference on Machine
Learning will be held. The purpose of this call is to invite groups
interested in hosting the conference to submit proposals. The group
selected to host the conference will work with a Program Chair and
Conference Chair chosen by the Board of the International Machine
Learning Society (IMLS).
Proposals should address the following issues:
1. Proposed Dates. The conference should be scheduled for four days
(one day reserved for workshops and tutorials; three days for paper
sessions, poster sessions, and invited talks). The conference
prefers dates in the range from June 15 to July 15, but other dates
will be considered, particularly if they permit co-location with
other conferences.
2. Locale Parameters.
- Accessibility. Is it easy and inexpensive for people (especially
graduate students) to travel to the conference site? (Compute
mean airfares from Europe, North America, and Asia. Include
ground transportation from relevant airports to the site.)
- Meeting Rooms, AV Equiment, etc. What are the physical
facilities like? Consider rooms for plenary sessions, parallel
sessions, workshops, tutorials, and poster sessions. What are
the charges, if any, for using them?
- Meals and Lodging. Is there low-cost, quality housing available
for attendees (especially graduate students)? How far from the
meeting rooms? Where will attendees eat? Please estimate costs
for meals and lodging.
- Demo facilities. Will there be computing equipment and space
available to support demos?
- Internet access. Is wired or wireless internet access available?
At what cost?
- Other features. You may mention any other aspects of the site or
the region that are relevant.
3. Local Machine Learning Community. Is there a local ML
group/community that can help with organization and funding?
4. Opportunities to co-locate with other conferences.
5. Organizational and Institutional Support. Is there a conference
office that can help with local arrangements?
Proposals (postscript or PDF) should be sent before November 1, 2003
to:
Tom Dietterich, President IMLS
[EMAIL PROTECTED]
Proposals will be ranked according to cost and accessibility, proposed
dates, opportunities for co-location, attractiveness of the location,
and experience of the host group. Preference will be given to
locations outside North America in accordance with the ICML policy of
siting the conference outside North America every third year.
------------------------------
End of ML-LIST Digest Vol 15, No. 18
************************************
